Genioglossus Muscle Is the Largest Extrinsic Tongue Muscle and Upper Airway Dilator

Home , Chondroglossus, Genioglossus, Geniohyoid muscle, Hyoglossus, Levator veli palatini, Masseter muscle

PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Kwan

First name: Benjamin Other name/s: Chi Hin

Abbreviation for degree as given in the University calendar: PhD

School: Prince of Wales Hospital Clinical School Faculty: Medicine

Title: Breathing movements of the human tongue and genioglossus measured with ultrasound imaging

Abstract 350 words maximum: (PLEASE TYPE)

Genioglossus muscle is the largest extrinsic tongue muscle and upper airway dilator. To maintain pharyngeal patency within and between breaths, delicate moment-to-moment coordination of pharyngeal muscles activity and drive is required. Dynamic pharyngeal muscle movement in response to the neural input during sleep/wake states is not clearly understood. This thesis reports a novel ultrasound method to visualise and measure dynamic genioglossus motion in healthy and OSA subjects. In Chapter 2, the method revealed ~1 mm predominantly anterior peak displacement within a 50 mm2 area in the infero-posterior genioglossus in healthy awake subjects during quiet breathing. Motion within this area was non-uniform. The method has good reliability, intraclass correlation coefficient (ICC) of 0.85 across separate imaging sessions. Chapter 3 reported good agreement between ultrasound and tagged MRI in measuring regional tongue motion in healthy and OSA subjects, with an ICC of 0.79. Compared to MRI, ultrasound revealed greater anterior displacement in the posterior tongue (mean difference of 0.24 ± 0.64 mm, 95% limits of agreement: 1.03 to -1.49). Chapter 4 examined influence of respiratory mechanics and drive on genioglossus movement. Inspiration against a resistive load increased posterior genioglossus motion, but it had less anterior and more inferior displacement at the highest inspiratory resistance. An increase in voluntary respiratory drive resulted in increased motion and a more uniform motion within the infero-posterior genioglossus. No significant change in genioglossus motion was observed with alteration of end-expiratory lung volumes. Chapter 5 examined the patterns of posterior tongue motion in awake healthy and OSA subjects. More uniform peak inspiratory motion was recorded within the posterior genioglossus in OSA subjects. Three types of inspiratory posterior tongue motion were observed, with breath-to-breath variability within and across subjects. There may be an association between tongue motion pattern with BMI and tongue-base angle. Likely effect of the motion is to counterbalance the negative pharyngeal collapsing forces. Variation between breaths and individual is possibly due to local anatomical, neural and biomechanical factors. Future research to investigate the biomechanical behaviour of the tongue in OSA subjects during sleep with concurrent neural drive measures may further our understanding into OSA pathogenesis.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

Signature Witness Signature Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………………………......

Date ……………………………………………......

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only). I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed ……………………………………………......

Date ……………………………………………......

AUTHENTICITY STATEMENT

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signed ……………………………………………......

Date ……………………………………………......

Breathing movements of the human tongue and genioglossus measured with ultrasound imaging

Benjamin Chi Hin Kwan

A thesis in fulfilment of the requirements for the degree of

Doctor of Philosophy

Neuroscience Research Australia

Faculty of Medicine, Prince of Wales Hospital Clinical School

University of New South Wales

August 2018

Abstract

Acknowledgements

First and foremost, I would like to express my sincere gratitude to Prof. Simon Gandevia for the continuous support of my PhD study, for his tremendous patience, motivation, guidance, understanding and immense knowledge. His mentorship and guidance helped me in all the time of research and writing of this thesis. His selfless time and care encourages me to become a better mentor and teacher to my own students and junior team members. I could not have imagined having a better advisor and mentor for my PhD study.

I would also like to thank Prof. Lynne Bilston for her insightful comments, knowledge, patience, support and encouragement. Amongst many things, she especially taught me the importance of attention to detail, and also for the hard questions which challenged and incented me to widen my research and reading from various perspectives.

I would also like to thank my other supervisor, A/Prof Jane Butler for her support, knowledge, and encouragement. Her positive attitudes are like an eternal beam of sunshine, and much appreciated during these years.

I sincerely thank Prof Rob Herbert, A/Prof David McKenzie, Dr Anna Hudson, Dr Shaokoon Cheng, Dr Julian Saboisky, Dr Billy Luu, Dr Lauriane Jugé, Dr Peter Burke, Dr Martin Heroux, Fiona Knapman, and Alice Hatt for their help in teaching me how to use the various equipment for my studies, how to use different statistics program (and which statistics to run), the technical aspects of analogue signals, how to use computer programs for analysis of the MRI and ultrasound images, helping with equipment repair and importantly sit through multiple experiments with me to aid me in collecting data. Their willingness to help me, sometimes at short notice, demonstrates the excellent camaraderie at NeuRA. I am also grateful for the staff of the NeuRA workshop for constructing the head-out rigid-shell ventilator, and the staff of the NeuRA IT department for forever helping me with my aging computer (circa 2010) and granting me access to the uploaded ultrasound files. I am also most grateful to all the subjects for their time and patience, especially those who participated in repeated studies due to equipment failure or computer program error.

iii

I especially want to thank Ben Tong for keeping me well fed, energised and motivated on research days with food, coffee and laughs. He is also invaluable in assisting me during many of my experiments.

I would like to express a special gratitude to my other mentor and friend, A/Prof Tara MacKenzie, for her support, encouragement, knowledge and insights through the years.

I also need to pass a big thank you to my colleagues at the Sutherland Hospital, as well as Gideon, Chloris, Megan, Mackie, Antonia, Helen Stavrou, Nicky Bennie, and most of all Jennifer Simmons, for being so supportive, accommodating, and encouraging all these years. And for making me laugh along the way.

Thank you also to the National Health and Medical Research Council of Australia and NeuRA for funding part of this research, and to the ResMed Foundations scholarship (awarded by the Royal Australasian College of Physicians) for providing me much needed support in the early years of my PhD journey.

I also need to thank my friends who provide me with so much support and encouragements over the years, sharing with me their experiences, their journeys and also being there when I need to chat or debrief.

I also like to express a sincere gratitude to my family, my sister Alisa and her family (Hiep and Aurelia), and especially my mother Anna. Thank you for being there, for supporting me and my family during these times, especially mum, caring for my family and I when we are all sick or when we need that extra support, no matter how late we call. I also like to thank my in-laws, Len and Denise, for providing me much needed support and encouragements.

Finally, I cannot thank my husband Matthew enough for his continuing love, support, sacrifices and understanding through these years of me pursuing my PhD degree while juggling a busy clinical load. This journey would not have been possible without your unfailing support and encouragement. I also thank our beautiful children, Jacob and Emily, whose arrival during this journey make it all the more special. Their kisses, hugs and smiles are the best source of encouragement especially during the most stressful days. Thank you all of you for being there for me.

Papers and conference proceedings resulting from this thesis

Kwan, B., Jugé, L., Bilston, L., Gandevia, S., 2017. Sagittal measurement of genioglossus movement during respiration: comparison between ultrasonography and magnetic resonance imaging. Sleep Medicine 40, e175.

Unpublished manuscripts submitted for publication

Kwan, B., Jugé, L., Bilston, L., Gandevia, S. Sagittal measurement of tongue movement during respiration: comparison between ultrasonography and magnetic resonance imaging. Manuscript sumbitted for publication.

Publications

Kwan, B.C., Butler, J.E., Hudson, A.L., McKenzie, D.K., Bilston, L.E., Gandevia, S.C., 2014. A novel ultrasound technique to measure genioglossus movement in vivo. J Appl Physiol 117, 556-562.

Kwan, B.C., Butler, J.E., Hudson, A.L., McKenzie, D.K., Bilston, L.E., Gandevia, S.C., 2015. Erratum: a novel ultrasound technique to measure genioglossus movement in vivo. J Appl Physiol 118, 1330.

Kwan, B.C.H., McBain, R.A., Luu, B.L., Butler, J.E., Bilston, L.E., Gandevia, S.C., 2018. Influence of respiratory mechanics and drive on genioglossus movement under ultrasound imaging. PLoS One 13, e0195884.

Table of contents

Abstract………………………………………………...... ii

Acknowledgements…………………………………………………….…………...iii

Publications and conference proceedings resulting from this thesis………………..v

Table of contents……………………………………………………………………vi

List of Figures and Videos…………………………………………...... xi

List of Tables……………………………………………………...... xiii

List of abbreviations………………………………………………….……...... xv

LITERATURE REVIEW

CHAPTER 1 Introduction and literature review………………………… 2

1. Anatomy and action of human upper airway muscles

1.1 Anatomy and action of upper airway respiratory muscles……………………...3

1.2 Extrinsic muscles of the tongue………………………………………………... 4

1.3 Intrinsic muscles of the tongue………………………………………………… 6

1.4 Other upper airway muscles – suprahyoid muscles……………………………. 8

1.5 Other upper airway muscles – infrahyoid muscles………………………...... 8

1.6 Other upper airway muscles – muscles of the soft palate…………………….... 9

1.7 Other upper airway muscles – pharyngeal constrictor muscles………………... 9

1.8 Fibre composition of the human tongue……………………………………….. 10

2. Neural control of upper airway muscles

2.1 Cortical drive to upper airway muscles…………………………………………13

2.2 Neural control of the tongue…………………………………………………….16

2.3 Neurotransmitters and pharyngeal motor control…………………………....….17

3. Upper airway muscles during sleep in healthy and obstructive sleep apnoea

3.1 Normal sleep wake cycle…………………………………………………....…..19

3.2 Neural control of sleep……………………………………………………....…..20

3.3 Basic respiratory neurobiology during sleep………………………………....….20

3.4 Behaviour of the upper airway during sleep…………………………………..…21

3.5 Upper airway in obstructive sleep apnoea…………………………………...…..22

3.6 Impact of aging on the upper airway…………………………………...………..28

4. Imaging of upper airway in normal and OSA subjects

4.1 Imaging modalities of the upper airway………………………………………….29

4.2 Upper airway imaging during sleep/wake states…………………………………36

5. Ultrasound imaging of the upper airway

5.1 Diagnostic ultrasound physics……………………………………………………39

5.2 Interaction with tissues by ultrasound waves…………………………………….39

5.3 Diagnostic ultrasound equipment………………………………………………...42

5.4 Image artifacts…………………………………………………………………... 42

5.5 Ultrasound image quality…………………………………...……………………43

5.6 Ultrasound of the airway…..……………………………………………………..44

5.7 Ultrasound image of the tongue….…………...………………………………….46

5.8 Image correlation technique and its application……………………………….…47 Outline of subsequent Chapters……...... …………………………………...49

EXPERIMENTS

CHAPTER 2 A novel ultrasound technique to measure genioglossus movement in vivo……………………………………………………………..52

1. Introduction………………………………………………………………………...... 53

2. Methods……………………………………………………………………………………54

2.1. Experiment protocol………………………………………….……………….…54

2.2. Ultrasound scanning and analysis………………………………….……………56

vii

2.3. Repeatability…………………………………………………………………….57

2.4. Statistical analysis……………………………………………………………….58 3. Results……………………………………………………………………………………..59 3.1 Visualisation of oral anatomy………………………………………………..…..59

3.2 Displacement during quiet breathing …………………………………….……...60

3.3 Timing of inspiratory displacement……………………………………………...64

3.4. Repeatability of ultrasound measure…………………………………………….64

4. Discussion…………………………………………………………………………………64 4.1. Genioglossus displacement during quiet breathing…………...... …………...65 4.2. Application of ultrasound method………………………...... ………………….66 4.3. Limitations……………………………………………………………...……….66 5. Conclusion………………………………………………………………………………....67

CHAPTER 3 Sagittal measurement of tongue movement during respiration: comparison between ultrasonography and magnetic resonance imaging ………………………………………………………………………..68

1. Introduction………………………………………………………………………...... 69

2. Methods……………………………………………………………………………………70

2.1. Experiment protocol………………………………………….……………….…71

2.2. Ultrasound scanning and analysis………………………………….……………72

2.3. MRI scanning and analysis ……………………………………………………..73

2.4. Comparing regional movement during inspiration between MRI and ultrasound75

2.5. Statistical analysis……………………………………………………………….75 3. Results……………………………………………………………………………………..76 3.1. Anterior grid……………………………………………………………………..79 3.2. Posterior grid……...……………………………………………………………..79 3.3. Posterior column of 5 points…………………………………………………….82 4. Discussion…………………………………………………………………………………85 4.1. MRI applications and limitations………………………………………………..86

viii

4.2. Ultrasonography applications and limitations…………………………………...87 5. Conclusion………………………………………………………………………………....88

CHAPTER 4 Influence of respiratory mechanics and drive on genioglossus movement under ultrasound imaging……………………………………….89

1. Introduction………………………………………………………………………...... 90

2. Methods……………………………………………………………………………………91

2.1. Experiment protocol………………………………………….……………….…92

2.2. Ultrasound scanning and analysis………………………………….……………93

2.3. Genioglossus motion during voluntary hyperpnoea…………………………….95

2.4. Genioglossus motion with inspiratory resistive loading………………………...95

2.5. Genioglossus motion with imposed changes in end-expiratory lung volume…..96

2.6. Statistical analysis………………………………………………………………96

3. Results…………………………………………………………………………………….97

3.1. Voluntary hyperpnoea…………………………………………………………..98

3.2. Inspiratory resistive loading……………………………………………………101

3.3. Change in lung volume produced by an external ventilator……………………104

4. Discussion………………………………………………………………………………..106

4.1. Limitations……………………………………………………………...……...109 5. Conclusion……………………………………………………………………………….110

6. Supplementary tables…………………………………………………………………….111

CHAPTER 5 Tongue movement during respiration in obstructive sleep apnoea under ultrasound imaging………………………………………….115

1. Introduction………………………………………………………………………...... 116

2. Methods…………………………………………………………………………………..118

2.1. Experiment protocol………………………………………….………………...119

2.2. Image analysis………………….………………………………….…………...119

2.3. Statistical analysis……………………………….……………………………. 121

3. Results……………………………………………………………………………………122

3.1. Anterior grid……………………………………………………………………123 3.2. Posterior grid…………………………………………………………………...124 3.3. Posterior column and tongue movement pattern……………………………….125 4. Discussion………………………………………………………………………………..130 4.1. Regional tongue displacement during inspiration……………………………...131 4.2. Effect of OSA risk factors on pattern of posterior tongue motion……………..132 4.3. Limitations……………………………………………………………………...134 4.4. Implications…………………………………………………………………….135 5. Supplementary table..…………………………………………………………………….136

CONCLUSION

CHAPTER 6 Summary of key findings and conclusions…..……………..137

1. General discussion………….…………………………………………...... …...138

1.1. Dynamic tongue motion measured with ultrasonography…………………...... 138

1.2. Influence of respiratory drive and mechanics on tongue motion………………141

1.3. Implications of the pattern of tongue motion in OSA subjects………………...142

1.4. Limitations of the ultrasound method and future directions…………………...145

2. Conclusions……………………………………………………………………………...147

References……………………………………………………………………149

List of Figures and Videos

Figure 1.1. 3D schematic models of the human tongue musculature……………….….. 6

Figure 1.2. Schematic diagram of the dorsal brainstem and cervical spinal cord………14

Figure 1.3. Proposed schematic overview of central respiratory pattern generator model in humans………………………………………………………………………………………15

Figure 1.4. Schematic model illustrating interplay between upper airway patency, upper airway muscle activities and luminal pressure………………………………………………24

Figure 1.5. Illustration of current upper airway biomechanical concepts………………..25

Figure 1.6. Tracked movement of a grid of points on the tongue during a respiratory cycle in a single subject during a MR tagging study………………………………………………35

Figure 1.7. Schematic of tissue interactions with ultrasound wave……………………...40

Figure 2.1. Schematic representation of ultrasound image, grid and numbering system, and subject positioning………………………………………………………………………55

Figure 2.2. Displacement of the infero-posterior region of genioglossus in a typical subject over 3 imaging sessions………………………………………………………...……60

Figure 2.3. Mean displacement of the infero-posterior group within posterior genioglossus grid for 20 subjects over a typical breath……………………………………………...... …..62

Figure 2.4. Resultant displacement of all points on the grid for genioglossus………...... 63

Figure 3.1. Set up for ultrasonography and MRI studies including grid positions……….72

Figure 3.2. Sites of major anatomical measurement on MRI images…………………….74

Figure 3.3. Bland-Altman plots of the difference between ultrasound and magnetic resonance imaging in measuring tongue displacement………………………………………80

Figure 3.4. Scatter plots of the tongue grids and posterior column displacement between ultrasound and magnetic resonance imaging…………………………………………………81

Figure 3.5. Scatter plots of the posterior tongue regional displacement between ultrasound and magnetic resonance imaging…………………………………………………………….82

Figure 3.6. Peak inspiratory movement of anterior and posterior tongue grids and posterior column measured with MRI and ultrasound for one representative subject……….83

Figure 4.1. Setup for each of the 3 conditions. Ultrasound transducer is positioned submentally…………………………………………………………………………………..93

Figure 4.2. Schematic representation of ultrasound image, grid position and numbering system……………………………………………………………………………………… ..94

Figure 4.3. Mean peak inspiratory movement of grid during voluntary hyperpnoea experiment…………………………………………………………………………………..101

Figure 4.4. Mean peak inspiratory movement of grid for 20 subjects during resistive inspiratory load experiment and lung volume alteration experiment……………………….103

Figure 4.5. Mean peak inspiratory movement of grid for 6 subjects inside or outside chamber during lung volume alteration experiment………………………………………...105

Figure 5.1. Locations for tracking grids and posterior column and anatomical measurements……………………………………………………………………………….120

Figure 5.2. Average maximal inspiratory movement of tracking grid in anterior and posterior tongue across healthy and OSA subjects…………………………………………124

Figure 5.3. Posterior tongue movement patterns and observed pattern distribution across different OSA severity……………………………………………………………………...126

Figure 5.4. Posterior tongue movement patterns in 4 representative subjects………….127

Supplementary Videos embedded within the accompanied Powerpoint presentation.

Video S.1. An ultrasound recording of tongue motion in a typical subject.

Video S.2. An ultrasound and tagged MRI recording of tongue motion of the same subject.

Video S.3. Tracking of infero-posterior tongue region (posterior grid) on ultrasound and MRI of the same subject during one respiratory cycle.

xii

List of Tables

Table 1.1. Common values for soundwave propagation velocity and attenuation in different medium/tissue………………………………………………..……………….…….41

Table 1.2. Common values for attenuation coefficient and penetration depth.……….....41

Table 1.3. Percentage of incident energy reflected at tissue interfaces assuming flat interface and perpendicular incidence………………………………………………………..45

Table 2.1. Characteristics of 20 subjects………………………………………………...59

Table 2.2. Average displacement of defined sub-groups in the posterior genioglossus...61

Table 2.3. Average displacement of each grid points for 20 subjects…………………...63

Table 3.1. Characteristics of 21 subjects………………………………………………...76

Table 3.2. Image grid / column characteristics across experiments……………………..77

Table 3.3. Reproducibility of measurements…………………………………………….78

Table 3.4. Maximal inspiratory displacement of 5 posterior column points…………….84

Table 3.5. Variability of measured mean maximal inspiratory regional tongue displacement………………………………………………………………………………….85

Table 4.1. Characteristics of the total pool of 38 subjects……………………………….97

Table 4.2. Image grid characteristics across experiments……………………………….98

Table 4.3. Respiratory variables for the voluntary hyperpnoea experiment…………….99

Table 4.4. Average displacement of infero-posterior region of genioglossus (voluntary hyperpnoea)…………………………………………………………………………………100

Table 4.5. Respiratory variables for the inspiratory resistive load experiment………...102

Table 4.6. Average displacement of infero-posterior region of genioglossus (inspiratory resistive load)……………………………………………………………………………….103

Table 4.7. Average displacement of genioglossus (lung volume alteration)…………..105

xiii

Table 4.8. Summary of genioglossus activity during different respiratory states……...107

Table 4.S.1. Mean maximal inspiratory displacement of 15 grid points during voluntary hyperpnoea experiment for 20 subjects……………………………………………………..111

Table 4.S.2. Mean maximal inspiratory displacement of 15 grid points during inspiratory resistive load experiment for 20 subjects…………………………………………………...112

Table 4.S.3. Mean maximal inspiratory displacement of 15 grid points during lung volume alteration experiment for 20 subjects……………………………………………………….113

Table 4.S.4. Respiratory variables for the lung volume alteration experiment…………..114

Table 5.1 Characteristics of the 46 subjects and their OSA severity…...... …………...122

Table 5.2 Tongue grid / column characteristics……………………………………….123

Table 5.3 Peak inspiratory tongue motion of different groups within the anterior grid.124

Table 5.4 Peak inspiratory tongue motion of different groups within the posterior grid………………………………………………………………………………………….125

Table 5.5 Input variable’s unique contributions in the multinomial logistic regression (n=138)……………………………………………………………………………………...127

Table 5.6 Multinomial logistic regression analysis for posterior tongue movement pattern and input variables………………………………………………………………….128

Table 5.7 Input variable’s unique contributions in the multinomial logistic regression (n=138)……………………………………………………………………………………...129

Table 5.8 Multinomial logistic regression analysis for OSA severity and input variables…………………………………………………………………………………….130

Table 5.S.1. Frequency and row percentages table reporting association between OSA severity and observed tongue movement pattern…………………………………………...136

xiv

List of abbreviations

2D – 2-dimension

3D – 3-dimension

4D – 4-dimension

5-HT – 5-hydroxytryptamine

AHI – Apnoea/hypopnoea index

BMI – body mass index

CT – Computed tomography

CPAP – Continuous positive airway pressure

EEG – Electroencephalogram

EELV – End-expiratory lung volume

EMG – Electromyography

GG – Genioglossus

GH – Geniohyoid

ICC – Intraclass correlation coefficient

MAS – Mandibular advancement splints

MRE – Magnetic resonance elastography

MRI – Magnetic resonance imaging

MyHC – Myosin heavy chain

NREM – Non-rapid eye movement

OCT – Optical coherence tomography

OSA – Obstructive sleep apnoea

REM – Rapid eye movement

SD – Standard deviations

SMU – Single motor unit

SPAMM – Spatial modulation of magnetisation method tMRI – tagged MRI

UA – Upper airway

UPPP – Uvulopalatopharyngoplasty

US – ultrasound

xvi

LITERATURE REVIEW

CHAPTER 1

Introduction and literature review

SUMMARY

The upper airway is an important conduit for respiration, vocalisation and deglutition in humans. The pharyngeal airway lacks structural cartilaginous or bony support and is highly susceptible to collapse. Patency of the upper airway depends on structural and neuromuscular factors, in particular the balance between collapsing forces (negative intraluminal inspiratory pressure) and dilatory forces (upper airway dilatory muscles). The genioglossus muscle is the largest upper airway dilator. This balance is important during sleep, when tonic and phasic upper airway activities are altered with a reduction in upper airway muscles activities, associated with reduced negative pressure reflex response. Unfavorable interaction between an individual’s upper airway anatomy and sleep-related changes in upper airway behaviour will destabilise the system and result in loss of pharyngeal patency and development of sleep disordered breathing syndromes such as obstructive sleep apnoea. To further understand the pathogenesis of obstructive sleep apnoea and discover potential new treatment options, it is important to establish any mechanical consequence to upper airway muscles associated with altered respiratory mechanics during wakefulness and sleep in healthy and patients with obstructive sleep apnoea. This requires an imaging modality that ideally would be inexpensive, noninvasive, free of radiation, dynamic and provides real-time imaging to a subject in a supine position during sleep.

In this introduction, I will briefly summarise the anatomy and function of the upper airway muscles, and discuss the respiratory neural drives they receive, during wakefulness and sleep. In addition, a brief review of obstructive sleep apnoea and current concepts of its pathogenesis including upper airway mechanics / behaviour is presented. Finally, a summary of upper airway imaging modalities is given. In this Chapter, key reviews are usually cited, with inclusion of some examples of major papers in the field.

1. ANATOMY AND ACTION OF HUMAN UPPER AIRWAY MUSCLES

1.1 Anatomy and action of upper airway respiratory muscles

The upper airway (UA) muscles are important for numerous complex motor tasks including swallowing, speech articulation, mastication and most importantly, respiration. These muscles include the tongue and soft palate. During these motor tasks, the tongue provides support and patency to the pharyngeal space. Unlike other animals, evolution of hyoid bone in humans resulted in an anatomical free floating bone, generating a speech apparatus (Weissengruber et al., 2002). However, this come at the expense of a stable upper airway which is more collapsible and results in upper airway disorders such as obstructive sleep apnoea (OSA) (DeGusta et al., 1999; Kay et al., 1998). Previous imaging studies have demonstrated the velopharynx as the site most commonly associated with airflow limitation (e.g.Ciscar et al., 2001). Although the human tongue can move in a variety of directions; the major movements are protrusion and retraction. The former action assists in dilating the pharyngeal airway while the latter occludes it. However, it has been demonstrated in rodent studies that airway collapsibility is counteracted by co-activation of protruder and retractor muscles (Fuller et al., 1999).

In humans, the tongue muscle lacks tendon or bony attachments. It encompasses 4 pairs of muscles which originate outside the tongue but insert into the tongue muscle body – genioglossus, styloglossus, hyoglossus and palatoglossus; and 4 pairs that originate and insert into tongue muscle entirely – superior longitudinal, inferior longitudinal, vertical and transverse muscles. These are termed extrinsic and intrinsic muscles respectively (Sonntag, 1925). Animal experiments have suggested these two groups of muscles are arranged both in-series (Hellstrand, 1980) and in-parallel (McClung and Goldberg, 2000). A recent study reported the genioglossus, transversalis, verticalis and superior longitudinal muscles together composed 70% of both the functional and structural volume of the tongue, with genioglossus as the largest muscle (>20%) (Stone et al., 2016). They are involved in all 3 directions of tongue expansion and compression. Although the co-ordinated actions of these separate muscles’ individual fibres determine the action of the tongue and also its shape, the intrinsic tongue muscles act to alter tongue shape whereas the extrinsic tongue muscles alter the position of the tongue (e.g. Altschuler et al., 1994; Dobbins and Feldman, 1995; Sauerland

and Mitchell, 1975). Figure 1.1 is a combined schematic and 3D model of the human tongue musculature indicating the geometric arrangement of intrinsic and extrinsic tongue muscles (Sanders and Mu, 2013; Takemoto, 2001).

Through studying various different animal tongues and muscle groups, Kier & Smith described mammalian tongues to act as a “muscular hydrostat”, defining it to have a constant- volume, with change in tongue body shape determining movement (e.g. Gilbert et al., 2007; Kier and Smith, 1985; Smith and Kier, 1989). Furthermore, selective activation of different muscle groups can result in rotation or twist motion, and simultaneous activation of longitudinally and transversely orientated muscle fibres can result in stiffening of the tongue (Sanders and Mu, 2013). This allows the tongue to be able to change shape simultaneously in multiple dimensions through coordinated interaction between the intrinsic and extrinsic muscles. However, because the posterior tongue is extensively joined with other pharyngeal structures such as hyoid bone, cranial base and pharyngeal muscles unlike the anterior free end, the hydrostat mechanism and contractile pattern may be different between the two regions (e.g. Gilbert et al., 2007; Zaidi et al., 2013).

In vivo (Bennett and Hutchinson, 1946) and in vitro (Doran, 1975; McClung and Goldberg, 2000) mammalian tissue studies described tongue protrusion as a combined effect of intrinsic (verticalis and transversus) and extrinsic (genioglossus) muscles contraction. While tongue retraction involves the extrinsic styloglossus and hyoglossus muscles as well as intrinsic superior and inferior longitudinal muscles. These were also reported in frogs during feeding (Nishikawa et al., 1999) and in human lingual functions not related to respiration (e.g. swallowing, chewing) (Napadow et al., 1999a; Sauerland and Mitchell, 1975). The potential of the intrinsic muscles to affect the mechanical actions of the extrinsic tongue muscles has been reported in mammals, and they are likely to contribute significantly to lingual stiffness.

1.2 Extrinsic muscles of the tongue

In this group, genioglossus is the largest in size (Fig. 1.1A & I). When compared with other mammals, human genioglossus is also larger proportionally (Lowe, 1980). Its fibres spread radially, antero-posteriorly and medio-laterally from the superior mental spine of the mandible (Abd-El-Malek, 1938; Mu and Sanders, 2000; Saigusa et al., 2001; Silverstein et al., 2000) with its muscle fascicles align at different angles (Sanders and Mu, 2013). The

anterior part of genioglossus is thin with vertically oblique orientation of its muscle fascicles, with a separate horizontally orientated muscular compartment posteriorly, separated by a thin layer of connective tissue (Doran and Baggett, 1972). Different branches of the hypoglossal nerve innervate different parts of the genioglossus, and may suggest the presence of in-series arrangement of separate muscle fibres within genioglossus (Mu and Sanders, 2010; Sanders and Mu, 2013). It protrudes the tongue in frogs, rats and humans (Cunningham and Basmajian, 1969; Nishikawa et al., 1999; Pittman and Bailey, 2009; Travers and Jackson, 1992). In humans, genioglossus is active in speech, swallowing and respiration. It also pulls the hyoid and base of tongue antero-inferiorly during an isolated contraction to open the pharynx, mainly through the action of the horizontal genioglossus and posterior fascicles of the oblique genioglossus (Mu and Sanders, 2010; Sanders and Mu, 2013; Smith et al., 2005). Its other actions in humans include depression of the tongue body, inferior curving of and possibly retrusion of the tip of the tongue.

With its origin in the hyoid bone, the hyoglossus muscle (Fig. 1.1B & K) inserts its fibres into the styloglossus and inferior longitudinal muscle along the length of the tongue (Sanders and Mu, 2013; Takemoto, 2001), and its activation pulls the lateral edge of the tongue postero-inferiorly (Saigusa et al., 2004; Sanders and Mu, 2013). It is active during inspiration and is innervated by the lateral branch of the hypoglossal nerve. Its response to stimuli in humans and animals is similar to genioglossus (Fregosi and Fuller, 1997; Fuller et al., 1999; Mateika et al., 1999). Often described as a bundle of hyoglossus, the chondroglossus muscle originates from the hyoid bone and passes through inferior longitudinal muscle before inserting within genioglossus (Ogata et al., 2002).

Styloglossus muscle (Fig. 1.1G & J) is the most lateral tongue muscle spreads anteriorly and medially to the lateral border of the tongue. It inserts into the tongue posterior to the palatoglossal arch and originates from the anterior and lateral surfaces in the styloid processes of the temporal bone (Iskander and Sanders, 2003; Takemoto, 2001). Some of its fascicles can also originate from the stylomandibular ligament (Merida-Velasco et al., 2006). In humans and animals, electromyography (EMG) studies show that styloglossus contracts with genioglossus in response to hypoxia and hypercapnoea (Mateika et al., 1999; Yasui et al., 1993). Its activation elevates and retracts the lateral margin of the tongue (Sanders and Mu, 2013).

The smallest extrinsic muscle of the tongue is the pair of palatoglossus muscles (Fig.1.1G), which originate at the soft palatal aponeurosis on each side. They extend distally to the side of the tongue before joining the styloglossus (Takemoto, 2001). Active during inspiration, they lower the soft palate in unison to the posterior part of the tongue. This occludes the oropharynx and increases the velopharyngeal airway (Carlson et al., 1995; Mortimore and Douglas, 1996; Rodenstein and Stanescu, 1984; Series, 2002).

Figure 1.1 has been removed due to copyright restrictions. Figures are adapted from (Sanders and Mu, 2013; Takemoto, 2001).

Figure 1.1. 3D schematic models of the human tongue musculature.

3D models describing different structural units of human tongue muscle and interdigitating intrinsic and extrinsic tongue muscles fibres. (A), (B), (G) and (H) describe different layers from innermost to outermost musculature. (C), (D), (E), (I), (J) and (K) describe the specific intrinsic and extrinsic tongue muscles. (F) Lateral view of the tongue with different colours denote different tongue muscle. Legend for the muscles is shown at top of image. (L) Midsagittal anatomical model showing 3 specific muscles as marked by the colour in (F). (Adapted from Sanders and Mu, 2013; Takemoto, 2001)

GG – genioglossus muscle; T – transverse lingual muscle; V – vertical muscle; IL – inferior longitudinal muscle; HG – hyoglossus muscle; PG – palatoglossus muscle; SG – styloglossus muscle; SL – superior longitudinal muscle.

1.3 Intrinsic muscles of the tongue

The intrinsic tongue muscles lack bony attachments and form the main part of the body of the tongue. They have their origin and insertion within the tongue with their fibres arranged in parallel and perpendicular to the tongue’s long axis. They interdigitate with the extrinsic tongue muscles (Stal et al., 2003) and provide a mechanism to control the airway size and shape (Bailey et al., 2006). Compared to limb, orofacial and masticatory muscles, the

intrinsic muscles show inter- and intramuscular differences in muscle fibre types (predominantly type I, IIA and IM/IIC fibres) and composition between different regions of the tongue (high amount of type II fibres anteriorly where type I fibres predominate posteriorly). This most likely related to regional differences in function (Stal et al., 2003).

The superior longitudinal muscle (Fig.1.1C & H) has fibres that run anteroposteriorly, extending from tongue base to its tip, along the superior surface of the tongue, just below the mucous membrane (Abd-El-Malek, 1938; Sanders and Mu, 2013; Slaughter et al., 2005). A thin layer of connective tissue separates it from the transversalis. It shortens the tongue and elevates, assists in retraction of, or deviates the tip of the tongue through controlling specific groups of muscle fibres within the tongue body (Sokoloff, 2000).

The inferior longitudinal muscle (Fig. 1.1B & D) lines the sides of the tongue, and is joined to the styloglossus muscle (Abd-El-Malek, 1938; Takemoto, 2001). It originates in the hyoglossus laterally and genioglossus medially. It shortens and ventroflexes the tongue and assists in retraction of the tongue base in conjunction with the superior longitudinal muscle (Sanders and Mu, 2013).

The transverse lingual muscle (transversalis) (Fig. 1.1A & E) has fibres that extend from the mucous membrane on the sides of the tongue to the median septum, which divides the tongue in half sagittally. It runs below the superior longitudinal muscle and joins the superior pharyngeal constrictor muscle posteriorly (Abd-El-Malek, 1938; Saigusa et al., 2004; Stal et al., 2003; Takemoto, 2001). Simultaneous contractions of both cause narrowing of the tongue and constriction of the mid pharyngeal airway (Saigusa et al., 2004; Sanders and Mu, 2013).

The vertical muscle (verticalis) (Fig. 1.1A & E) has fibres that originate near the midline of the superior surface of the tongue in the superior longitudinal muscle. It courses inferolaterally (down and to the sides of the tongue) to insert into the tongue dorsum that consists of the genioglossus, inferior longitudinal, and styloglossus muscles (Stal et al., 2003; Takemoto, 2001). It supports and forms an interdigiting structure with the transverse muscle between the longitudinal muscles (Saito and Itoh, 2003), and its action flatten the tongue (Sanders and Mu, 2013).

1.4 Other upper airway muscles – suprahyoid muscles

The suprahyoid muscles attach the hyoid bone to the skull. The geniohyoid muscle originates from the inferior mental spine on the internal surface of the anterior mandible and terminates at the anterior border of the hyoid bone. It elevates the hyoid bone forward and upward, thus dilating the airway (Series, 2002; Takahashi et al., 2002). However, there is minimal phasic activity during quiet respiration (Brown et al., 2011).

The mylohyoid muscles are a paired thin, flat muscle running from the mandible to the hyoid bone under the genioglossus, forming the floor of the oral cavity. It elevates and protracts the hyoid bone and raises the floor of the mouth and base of the tongue (Lowe, 1980). The mylohyoid muscle separates the sublingual space from the submandibular space, which communicate in a lateral gap between the mylohyoid and geniohyoid muscles (Otonari-Yamamoto et al., 2010).

The digastric muscle consists of two independent muscular bellies joined by an intermediate rounded tendon. The posterior belly arises from the mastoid of the temporal bone while the anterior belly arises from a fossa on the inner side of the lower border of the mandible. The intermediate tendon loops through a connective tissue sling, which is attached to the hyoid bone at the junction of the lesser horn and the body (Van Eijden et al., 1997; Wang et al., 2004). When the muscle contracts, it elevates the hyoid bone. If the hyoid is being held in place by the infrahyoid muscles, it will tend to depress the mandible and open the mouth.

The stylohyoid muscle is a slender muscle, lying superior and anterior to the posterior belly of the digastric muscle. It arises from styloid process of the temporal bone and inserts into the lateral side of the hyoid bone (van Lunteren and Strohl, 1986). The intermediate tendon of the digastric muscle perforates the stylohyoid near its insertion (Van Eijden et al., 1997). During swallowing, it elevates and moves the hyoid bone posteriorly (Hudgel, 1992).

1.5 Other upper airway muscles – infrahyoid muscles

Infrahyoid muscle group attach the hyoid bone to the larynx and sternum and are referred to as the “strap” muscles. As a group they move the hyoid bone and larynx inferiorly (Hudgel, 1992). The omohyoid muscle consists of two bellies joined by an intermediate tendon. The inferior belly is flat and narrow, arising from the upper border of the scapula

(Kasapoglu and Dokuzlar, 2007) and ends in the superior belly. The superior belly begins at the intermediate tendon, passes almost vertically upward to insert into the lower border of the hyoid, lateral to sternohyoid insertion (Kasapoglu and Dokuzlar, 2007).

The sternothyroid muscle arises from the posterior surface of the manubrium sterni, located underneath the sternohyoid and inserts into the lateral aspect of the thyroid cartilage (Henry et al., 2008).

The sternohyoid muscle is a thin muscle that originates from the posterior border of the medial end of the clavicle and the manubrium sterni. It is the most medial and superficial of the infrahyoid muscles (Galloway et al., 1991). It depresses the hyoid.

The thyrohyoid muscle depresses the hyoid and elevates the larynx. It arises from the superior aspect of the thyroid cartilage and inserted into the lower border of the greater cornu of the hyoid (van Lunteren and Strohl, 1986).

1.6 Other upper airway muscles – muscles of the soft palate

Four muscles that form the posterior arch of the oral cavity together with the palatoglossus. The tensor veli palatini is a broad, thin muscle that is found antero-lateral to the levator veli palatini, which is a thick, rounded muscle. The tensor veli palatini tenses the soft palate and assists the levator veli palatini, which is the main elevator of the soft palate, during swallowing to occlude the nasopharynx. It also assists in opening of the auditory tube during swallowing or yawning to equalise air pressure in the tympanic cavity. The palatopharyngeus muscle is a small muscle that arises from the soft palate and after joining the stylopharyngeus, inserts into the thyroid cartilage. It pulls the pharynx upward during swallowing (Casey, 1983; Ezzat and El-Shenawy, 2015). The musculus uvulae is entirely within the uvula. It broadens and shortens the uvula and changes the posterior surface of the soft palate. All of these muscles converge with extensive interdigitating of their fibres on a common central tendon plate (Barsoumian et al., 1998; Huang et al., 1997).

1.7 Other upper airway muscles – pharyngeal constrictor muscles

The posterior pharyngeal wall is comprised of an outer layer made up of three circularly disposed constrictor muscles. They are the superior, middle and inferior pharyngeal constrictors. Each of the 3 muscles is made up of a right and left part and all originated from the pharyngeal skeleton and insert into a midline raphe posteriorly, with the lower

constrictors overlap the one above. There are recorded anatomical and physiological interactions occur between the inferior posterior human tongue and the superior constrictor muscles (Kokawa et al., 2006; Saigusa et al., 2004). Through this interaction, it is possible any tongue movement imparts an indirect mechanical drag which provides stability and stiffness to the pharyngeal wall and the tongue (Fregosi, 2008; Fuller et al., 1999; Kuna, 2000).

The superior pharyngeal constrictor is a 4-sided muscle, with several origins from the medial pterygoid plate, pterygomandibular raphe, mandible near the mylohyoid line and from side of the tongue. It constricts the nasopharynx, especially during swallowing to aid food passage into the oesophagus.

The medial pharyngeal constrictor originates from the hyoid and is shaped like a fan. Its role is similar to the other pharyngeal constrictors but also elevates the hyoid. It also functions as the back wall of the vocal tract.

The inferior pharyngeal constrictor originates from the thyroid and cricoid cartilage. Besides its role in swallowing and acting as the back wall of the vocal tract, it also constricts the pharynx during pharyngeal consonants formation.

1.8 Fibre composition of the human tongue

The tongue is made up of a complex muscle fibre matrix (Anderson, 1881) due to interdigitating intrinsic muscles with the extrinsic tongue muscles. Within this complex organisation of muscles, separate nerve branches supply smaller neuromuscular compartments, and likely allows independent activation and movement of highly localised parts of tongue muscles (Sanders et al., 2013). One concept is to group fibre types according to their geometric orientation (Sokoloff and Burkholder, 2012). There are regional differences in this composition with a higher concentration of muscle in the posterior regions with larger fibre diameter (Miller et al., 2002; Stal et al., 2003). Vertically orientated fibres are mainly contributed by the genioglossus and verticalis muscle and are present in both anterior and posterior portion of the tongue (Abd-El-Malek, 1939; Gaige et al., 2007). Posterior tongue is most likely composed of genioglossus fibres, with verticalis occupying the lateral peripheral portion bilaterally. Transverse fibres are predominantly made up of transversalis and styloglossus muscles (Du Brul, 1976; Gaige et al., 2007; Saito and Itoh, 2007). Longitudinal fibres have significant interdigitation between intrinsic and extrinsic

muscles and receive contribution by genioglossus, hyoglossus, inferior longitudinal, styloglossus, superior longitudinal and transversalis muscles. In rodent and feline model, they have both in-series and in-parallel arrangements (Hellstrand, 1980; McClung and Goldberg, 2000) with predominance of in-parallel fibres in the posterior tongue.

Compared to other primates, human tongue muscles contain significantly more slow twitch muscle fibres, which provides precise control of movement. The proportion of slow twitch muscle fibres is similar in both extrinsic and intrinsic tongue muscles, except verticalis which contained a significantly lesser amount (Sanders et al., 2013). Regional differences in proportion and diameter of fibre types are also reported in intrinsic tongue muscles (Sokoloff and Burkholder, 2012; Stal et al., 2003). Anterior tongue mostly is made up of type II fibres (71%) which contains fast myosin heavy chain (MyHC) content. Posterior tongue has predominance of type I (slow MyHC content) and type IM/IIC fibres (mixed MyHC content) (Granberg et al., 2010; Sokoloff et al., 2010). There is also a higher proportion of slow twitch muscle fibres in the medial aspect of superficial longitudinal muscles compared to the lateral aspect (Sanders et al., 2013). There are further regional differences in distribution of capillary supply, with significantly higher capillary density in the posterior third of the tongue (Granberg et al., 2010), suggesting this region is more fatigue resistant (e.g. Crow and Ship, 1996; Zaidi et al., 2013). There are however differences in MyHC phenotype, as well as fibre compositions between intrinsic and extrinsic tongue muscles (e.g. Sokoloff et al., 2007). Genioglossus, the largest tongue muscle, is reported to have more type I fibres (51%) in posterior region compared to a predominance of type II fibres (68.7%) in anterior region (Saigusa et al., 2001), with a gradual increase in size of genioglossus fibres from anterior to posterior. The oblique portion of genioglossus also has more MyHC IIA and less MyHC I compared with horizontal portion (Daugherty et al., 2012). The change in muscle fibre size is independent of fibre type. This may suit the phasic actions seen in the anterior tongue region, compared with tonic or repetitive low-level contractions in posterior region (see Section 2. Neural control of upper airway muscles), particularly the biomechanical needs of the tongue to rapidly change its shape and movement during speech (Sanders et al., 2013).

In healthy controls, the amount of intramuscular connective tissue is also much higher in proportion in the human tongue (≈ 25%) compared to limb skeletal muscles (around 1-10%) (e.g. Kjaer, 2004; Humbert et al., 2008; Miller et al., 2002; for review, see Jarvinen et al., 2002). There is also an increase percentage of fat in human tongue compared to limb skeletal muscles, with a higher distribution in the posterior tongue region compared to anterior region

(Kim et al., 2014; Miller et al., 2002). A higher accumulation of fat was also reported in tongues of non-obese Zucker rats as compared to masseter muscle, another upper airway skeletal muscle (Brennick et al., 2014). The increased fat content may aid in the human tongue’s ability to maintain shape and structural strength (Rowe, 1981). A study in rodents (Brennick et al., 2008) and a cross-sectional human anatomical study (Nashi et al., 2009) further reported increased lingual fat content and weight with higher BMI. In the human autopsy study, the mean percentage of tongue fat is estimated to be 11% in the anterior tongue, 30% in the posterior tongue and 31% in the sublingual tissues.

2. NEURAL CONTROL OF UPPER AIRWAY MUSCLES

2.1 Cortical drive to upper airway muscles

The ventral respiratory column within the medulla contains the pacemaker cells that generate and regulate respiratory rhythm in mammals (Fig. 1.2). These are the preBötzinger complex (e.g. Feldman et al., 1990; Smith et al., 1991; Suzue, 1984) and the retrotrapezoid nucleus / parafacial respiratory group (e.g. Feldman et al., 2003; Onimaru and Homma, 2006). The former is believed to generate and regulate inspiratory rhythm while the latter controls expiration (e.g. Janczewski and Feldman, 2006; Mellen et al., 2003; for review, see Feldman et al., 2013) (Fig. 1.3). Further influence on human respiratory muscle activation has been recorded via cortical pathways from primary motor cortex (e.g. Colebatch et al., 1991; Gandevia and Rothwell, 1987; Murphy et al., 1990), the pre-motor cortex, and the supplementary motor area (e.g. Macefield and Gandevia, 1991; Sharshar et al., 2004). This convergence of multiple inputs at the motoneurones allows a flexible breathing pattern that enable essential tasks such as speech, swallowing and ventilatory rise with volitional tasks (e.g. Doty et al., 1967; for review, see Martin-Harris, 2006).

Pharyngeal airway muscles are innervated by several different nerves. The intrinsic and extrinsic tongue muscles are innervated by the hypoglossus nerve (O'Reilly and FitzGerald, 1990). The pharyngeal constrictor muscles receive their motor output from the pharyngeal branch of the vagus nerve and stylopharyngeus muscle is innervated by glossopharyngeal nerve, which also supply motor output via the pharyngeal plexus to levator veli palatini, pharyngeal constrictors, and cricopharyngeus muscles (Furusawa et al., 1991; Nishio et al., 1976). In breathing, the activity of hypoglossal motoneurones is shaped by both excitatory and inhibitory premotoneuronal influences. In mammals, the hypoglossal premotoneurones are primarily found lateral to the medullary hypoglossal motor nucleus (e.g. Holstege et al., 1977), but neuronal input and connections can also be found in the pontine parvocellular reticular formation (Rikard-Bell et al., 1985), the ventral medulla (Chamberlin et al., 2007), locus coeruleus and subcoeruleus (Aldes, 1990). Other input comes from the supratrigeminal, intertrigeminal and spinal trigeminal nuclei, as well as the ventrolateral medulla and the nucleus of the solitary tract in the dorsal respiratory group (Ono et al., 1994; Shintani et al., 2003). These respiratory premotor neurones exert significant influence on hypoglossal motoneurones (Orem and Kubin, 2000), for example hypoglossal respiratory

motor output can be activated or suppressed by neurones in the Kölliker-Fuse nucleus (e.g. Gestreau et al., 2005; Kuna and Remmers, 1999).

Fig. 1.2. Schematic diagram of the dorsal brainstem and cervical spinal cord.

Coloured regions are involved in respiratory rhythm. The colour palette represents the first-, second- and third-order neurons involved in diaphragm motor control. No specific caudal to rostral progression was reported. The approximate site of the hypoglossal motor nucleus is indicated on left (in yellow). (Adapted from Dobbins and Feldman, 1994; Saboisky, 2008)

Synchronisation of respiratory rhythm between the right and left hypoglossal nuclei is achieved through common excitatory connections from the pattern generators (Li et al., 2003b). During volitional behaviours (e.g. trunk rotation), the voluntary drive to inspiratory motoneurones may be modified directly via corticospinal projections or indirectly via the respiratory pacemaker neurones (e.g. Hudson et al., 2010, 2011; Nathan, 1963; for review, see Hilaire and Pasaro, 2003). There is also evidence that this drive is non-uniform across

pump muscles (Saboisky et al., 2007c) and that it can be preferentially directed to muscles offering the greatest mechanical advantage to produce inspiratory airflow (e.g. Gandevia et al., 2006; for review, see De Troyer et al., 2005). This model is termed “neuromechanical matching” and has been described in intercostal muscles, with the most advantageous motoneurones activated earlier and to a greater extent during a breath (for review, see Butler and Gandevia, 2008; Butler et al., 2014). This concept has not been described in upper airway muscles.

Figure 1.3 has been removed due to copyright restrictions. Figures are adapted from (Feldman et al., 2013).

Fig. 1.3. Proposed schematic overview of central respiratory pattern generator model in humans.

One proposed model describing two brainstem oscillators (preBötzinger complex and retrotrapezoid nucleus / parafacial respiratory group) that generate and regulate respiratory rhythms in humans. Red pathway described descending motor control to inspiratory muscles. Blue pathway described descending motor control to expiratory muscles. These can be modulated by neuromodulatory, suprapontine and sensory inputs. (Adapted from Feldman et al., 2013)

2.2 Neural control of the tongue

Predominant afferent and efferent control of the tongue is through the hypoglossal nerve (for review, see Mu and Sanders, 2010). Originating from the hypoglossal nucleus of the caudal brain stem, it exits the skull-base through the hypoglossal canal, follows the vagus nerve in close proximity passing between the internal carotid artery and internal jugular vein lying on the carotid sheath. It then passes deep to the posterior belly of the digastric muscle before passing between the mylohyoid and hypoglossus muscles where it then bifurcates into two branches (Lowe, 1980). The hyoglossus, styloglossus and inferior longitudinal muscles are innervated by the smaller lateral branch, while the larger medial branch innervates the genioglossus, transverse, vertical and superior longitudinal muscles (Lowe, 1980). The hypoglossal nerve innervates all extrinsic tongue muscles except the palatoglossus, which receives neural input from spinal accessory nerve (Salame et al., 2006).

In addition to the hypoglossal nerve, the lingual nerve also has direct anatomical connections in the tongue, and communicates with the hypoglossal nerves in the tongue, especially in the superior longitudinal muscle (Mu and Sanders, 2010). It is a branch of the mandibular division of the trigeminal nerve, but also carries fibres from the facial nerve. It supplies afferent sensory information from, and taste sensation to the anterior two thirds of the tongue. It also supplies the inferior longitudinal and superior longitudinal muscles (e.g. Fitzgerald and Law, 1958; Saigusa et al., 2006b; Zur et al., 2004).

In previous animal and human studies, the tongue muscles have different activation pattern depending on tasks (for review, see Lowe, 1980). During hypercapnoea in rats, extrinsic tongue muscles activate early (hyoglossus) before intrinsic muscles (superior longitudinal) (Bailey and Fregosi, 2004), which may be due to the functional sub-compartments of the hypoglossal nucleus and their various pre-motor inputs (e.g. Cunningham and Basmajian, 1969; for review, see Duffin, 2004; Sawczuk and Mosier, 2001). This is demonstrated in previous anatomical tracing experiments in which there are distinct distributions of dendrites in the hypoglossal nucleus for separate pharyngeal muscles (Dobbins and Feldman, 1995; Sokoloff and Deacon, 1992). Both common and separate innervation patterns are seen. In these models, the ventral compartment contains motoneurones for the protruders (e.g. Dobbins and Feldman, 1995; Gestreau et al., 2005; McClung and Goldberg, 1999) and the retractors cell bodies are located in the dorsal compartment (e.g. Gestreau et al., 2005). In contrast, genioglossus motoneurones are found in two separate pools in the ventral

compartment, with the lateral aspect controlling the horizontal fibres and vertical fibres controlled by the centrolateral nucleus (McClung and Goldberg, 2002). The latter also innervate the intrinsic tongue muscles. In rats, a small number of interneurons are also described (Popratiloff et al., 2001). This intricate subcompartmental motoneurones arrangement may provide the flexibility for the tongue muscles to control different movements associated in complex motor tasks such as speech, swallow and respiration (Altschuler et al., 1994). This can be demonstrated by a recent genioglossus single motor unit (SMU) and whole muscle EMG study which recorded activation of both genioglossus and intrinsic muscles during impeded and unimpeded tongue protrusion (Pittman and Bailey, 2009). The study supported the central role of genioglossus in tongue protrusion, but also the important role of the intrinsic muscles (in particular within the anterior tongue) in generating protrusion force. Activation of genioglossus during protrusion is likely to counterbalance the elongation of the tongue associated with activation of intrinsic muscles. Furthermore, in mammals, respiration is associated with co-activation of both tongue protruders and retractors (Fuller et al., 1998; Fuller et al., 1999; Oliven et al., 2007). This not only controls the position of the tongue, but also likely to increase its stiffness, reduce pharyngeal resistance and increase pharyngeal inspiratory flow (e.g. Eisele et al., 1997; Yoo and Durand, 2005). Another study of genioglossus EMG using a new long needle (with multiple) recording surfaces found suggestive evidence for larger motor unit territories in the anterior and inferior (above geniohyoid) regions of genioglossus compared with more posterior regions (Luu et al., 2017). Interestingly the units in the superficial anterior region discharged predominantly in expiration. This raises the possibility of regional activation within the genioglossus and further studies will be needed to assess this definitively.

2.3 Neurotransmitters and pharyngeal motor control

Previous studies have demonstrated EMG activity in upper airway muscles prior to inspiratory airflow or diaphragm activity during normal quiet breathing (e.g. Connel and Fregosi, 1993; Haxhiu et al., 1984; Onal et al., 1981; Saboisky et al., 2007c; Strohl et al., 1980; Van Lunteren et al., 1983). This is achieved via a coordinated descending central respiratory drive (e.g. Hwang et al., 1983b; Malhotra et al., 2000a; St-John et al., 2004; Tsuiki et al., 2000; Wheatley et al., 1991). The pharyngeal muscles, including genioglossus, typically show phasic activity during inspiration on a background of tonic activity that persists in expiration (e.g. Hwang et al., 1983a; Pillar et al., 2000b; Saboisky et al., 2006;

Sauerland and Harper, 1976; Sauerland and Mitchell, 1975; St-John et al., 2004; Tangel et al., 1992; for review, see Butler and Gandevia, 2008).

During sleep, this tonic genioglossus activity is suppressed the most (e.g. Pillar et al., 2000b; Sauerland and Harper, 1976; Tangel et al., 1992; for review, see Trinder et al., 2014). Experiments in intact (freely behaving) and reduced animal preparations suggest an important role for serotonergic (5-hydroxytryptamine (5-HT)) input modulation to hypoglossal motor neurones as cause of this suppression (e.g. Jelev et al., 2001; Kubin et al., 1998). The raphe nucleus is the only source of serotonergic input to the hypoglossal motoneurones (e.g. Manaker and Tischler, 1993; Woch et al., 1996). Medullary raphe neurones exhibit declining discharge rates during transition from wakefulness to sleep. Pharmacological model of rapid eye movement (REM) sleep in decerebrate cats (Kubin et al., 1994; Woch et al., 1996), and in freely behaving, naturally sleeping rats (Jelev et al., 2001) confirmed minimal raphe neuronal activity in REM sleep (Jacobs and Azmitia, 1992). In the rodent model, tonic genioglossus activation occurred when 5-HT was directly applied to the hypoglossal motor nucleus, with maintenance of this increased genioglossus muscle activity for several hours for as long as 5-HT was applied. This excitatory effect on genioglossus activity was modulated by the prevailing sleep/awake state, suggesting possible different neuronal mechanisms that impact on hypoglossal motoneurones in REM and non-REM sleep (NREM) (Horner, 2001).

This state-dependent modulation of pharyngeal muscle activity may also involve other neurotransmitters. For excitatory neurotransmitters, discharge rates of noradrenergic neurones of the locus coeruleus complex have been reported to decline from waking to non- REM sleep, with minimal activity in REM (Aston-Jones and Bloom, 1981), potentially contribute to decreases in the excitation of pharyngeal motoneurones during sleep. Although studies are lacking in human, in rat models, thyrotropin-releasing hormone and substance P, which co-localised with 5-HT, may possibly contribute to suppressed motor activity and reflex responses with their withdrawal (Bayliss et al., 1992; White, 1985). In feline models, inhibitory neurotransmitters such as γ-aminobutyric acid and glycine may promote suppression of hypoglossal motor output to genioglossus and thus its activity in REM (Fung et al., 2000; Kubin et al., 1993).

3. UPPER AIRWAY DURING SLEEP IN HEALTHY AND OBSTRUCTIVE SLEEP APNOEA

3.1 Normal sleep wake cycle

Sleep is a natural recurring complex behavioural state characterised by decreased response to the environment with voluntary muscles inhibition and relatively inhibited sensory activity. It is vital for survival in many species (e.g. Greenspan et al., 2001; for review, see Siegel, 2005). This sleep-wake cycle is important to many body systems including the immune, endocrine, nervous, skeletal, and muscular systems to ensure optimal performance (e.g. Everson and Toth, 2000; Orban et al., 2006; Peigneux et al., 2001; Rasch et al., 2007; Rauchs et al., 2008; Smith, 2001; Spiegel et al., 2000; Spiegel et al., 2004; Spiegel et al., 1999; for review, see Stickgold and Walker, 2007). It is still poorly understood as to the exact purposes and mechanisms of sleep.

During sleep, the body alternates approximately every 90 min between two distinct states: NREM and REM sleep (for review, see Jouvet, 1967). NREM sleep is further subdivided into three stages (1 through to 3) defined along electroencephalogram (EEG) and represents a deeper state of sleep continuum, with lowest arousal threshold in stage 1 (Williams et al., 1973). NREM sleep is typically associated with minimal or fragmentary neuronal activity, and other physical parameters such as brain temperature, breathing rate, heart rate and blood pressure are all at their lowest levels (Carskadon and Dement, 2001). Stage 1 NREM sleep is transitional stage between wakefulness and sleep, with slowing of EEG (4 – 7 Hz theta range) (Lavie et al., 2002). Typically, it represents only about 5% of the total sleep time. Stage 2 NREM sleep represents a complete loss of conscious awareness with emergence of two distinguishing phenomena, “sleep spindles” and “K-complexes” which may have a role in suppressing the response to external stimuli, as well as to refresh and consolidate memory (for review, see Colrain, 2005). This stage typically constitutes around 45-50% of total sleep time. As sleep deepens to NREM stage 3 sleep (slow wave sleep), low frequency delta wave (0.5 – 4 Hz) appears. Dreaming, memory consolidation and information processing takes place particularly during this stage. It tends to occur in longer periods during the first half of the night and occupies around 15-20% of total sleep time.

REM sleep occurs in cycles of 90 - 120 min and tends to occur in the latter part of the sleep period. It tends to reduce with age and occupies up to 20-25% of total sleep time in an

adult (Carskadon and Dement, 2001). It is associated with a high frequency, low amplitude cortical EEG signal, and phasic rapid eye movement (for review, see Fuller et al., 2006). Compared with NREM, breathing becomes more rapid and irregular, with increase in heart rate and blood pressure to near waking levels during REM sleep. Skeletal muscles atonia also occurs during REM (for review, see Williams et al., 1973).

3.2 Neural control of sleep

Multiple brain regions from neocortex to the lower brainstem have the capacity to general sleep and wake states in mammals (McGinty and Szymusiak, 2011). Several distinct neuronal groups synthesising and releasing chemicals such as acetylcholine, serotonin, norepinephrine, dopamine, histamine, orexin/hypocretin, and glutamate, have interspersed axons and terminals throughout the brain, and facilitate in wake and arousal (Moruzzi, 1972). The preoptic area within the hypothalamus is at the center of sleep promotion (Nauta, 1946; von Economo, 1930) and has reciprocal inhibitory connection with arousal-promoting systems, thus providing a balance in the activities between the two. A range of processes such as sensory, autonomic, endocrine, metabolic and behavioural influences can modulate both systems. Daily circadian rhythms are generated by the suprachiasmatic nucleus (Ralph et al., 1990).

3.3 Basic respiratory neurobiology during sleep

Postural muscle activity in mammals is highest during wakefulness, decreasing during NREM sleep and at its lowest during REM sleep, although there are occasional muscle twitches during REM associated with phasic REM sleep events (Horner, 2011). Respiratory muscle activity is similarly depressed (Worsnop et al., 1998). The diaphragm is the only exception. It has less suppression during NREM sleep and minimal motor inhibition during REM sleep. Overall, across sleep-wake states, respiratory muscle activity is modulated by non-respiratory (i.e. tonic) inputs (inhibitory or excitatory) from sleep-wake related regions in the brain, and from neurons activated by specific behavioral acts during wakefulness (Horner, 2008); as well as an additional rhythmic (central) respiratory drive (for review, see Horner et al., 2014). The influence of this non-respiratory tonic drive on respiratory motoneurons can have significant physiological effect, especially during reduced tonic drive from wakefulness to NREM sleep, and subsequent further reduction in REM sleep (Horner et al., 1994b; Orem

et al., 2000). One such effect is reduction in respiratory muscle activity and subsequent hypoventilation during NREM sleep.

During sleep, there is a further loss of the usual ventilatory response to hypoxic and hypercapnic drives, with the lowest responses occur during REM sleep (e.g. Berthon-Jones and Sullivan, 1982, 1984; Douglas et al., 1982a; Douglas et al., 1982b; Hedemark and Kronenberg, 1982; White, 1986; White et al., 1982; for review see Dempsey and Smith, 2014). Ventilatory response to added inspiratory resistance is also blunted during NREM sleep (e.g. Gugger et al., 1989; Hudgel et al., 1987; Iber et al., 1982; Wiegand et al., 1990).

3.4 Behaviour of upper airway during sleep

The upper airway is classically described as a Starling resistor, with the collapsible pharyngeal airway situated between the non-collapsible larynx and nasopharynx (e.g. Gleadhill et al., 1991; Schwartz et al., 1988; Smith et al., 1988). Flow through the pharyngeal airway is dependent on the opposing dilatory force (UA dilators and tracheal traction) and collapsing force (negative inspiratory transmural pressure and UA tissue pressure) (Heinzer and Series, 2011). The critical pressure (Pcrit) describes the pressure where UA closure occurs due to collapsing forces overcoming the dilatory forces (Ayappa and Rapoport, 2003). However, recent studies recorded individual variability in sites / structures where airflow limitation occur (Owens et al., 2014; Wellman et al., 2014), raising the possibility that different airflow limitation mechanisms occur across OSA patients.

During sleep, tonic and phasic UA activities are altered with a reduction in UA muscles activities (Horner et al., 1994a; Shea et al., 1999; Wheatley et al., 1993b). Different muscles have different activation profiles during sleep. In genioglossus and geniohyoid, the phasic inspiratory activity is maintained, but both tonic and phasic tone of genioglossus, geniohyoid, tensor palatini, levator palatini, palatoglossus and other UA muscles decreased at the onset of sleep (Ayappa and Rapoport, 2003). In normal subjects these decreased activities are short- lived, similar to EMG results recorded in the diaphragm and intercostal muscles (Tangel et al., 1992). Besides altered UA muscles activity, sleep may also alter the recruitment and activation pattern of UA muscles (e.g. Bailey et al., 2007; Hudgel and Harasick, 1990; Wilkinson et al., 2008, 2010), which may cause an increased UA resistance and lead to UA closure, thus causing instability of the UA flow model.

Other factors can also influence the stability of the UA. These include UA shape and cross-sectional area. Reduced lung volume increase UA collapsibility and reduce pharyngeal cross-sectional area (Burger et al., 1992a; Series et al., 1990). A change in the UA dimension such as UA axis shift from a transversal to an anteroposterior orientation; or reduction of UA cross-sectional area from start of expiration (maximal) to minimal at end of expiration during a respiratory cycle can also increase UA resistance and results in increased UA collapsibility (Leiter, 1996; Rodenstein et al., 1990). Upper airway patency can also be affected by its surrounding soft tissues which can affect tissue stiffness. Lengthening of the UA muscles during inspiration can also affect the action of the UA dilatory muscles (Brennick et al., 1997).

3.5 Upper airway in obstructive sleep apnoea

A range of upper airway dysfunctions can occur during sleep. These include increased upper airway resistance, snoring (due to turbulent airflow causing vibration of pharyngeal mucosal tissues), partial airflow limitation and airway closure (hypopnoea), and complete airway obstruction with cessation of airflow (apnoea). The diminished airflow in a hypopnoea or apnoea has to last for at least 10 sec and can persist to more than 60 sec (Young et al., 1993). Apnoeic events often result in cortical arousal and subsequent restoration of pharyngeal patency (Remmers et al., 1978). However, airway patency can be restored without cortical arousals in OSA subjects, most likely due to increased chemical drive (Younes, 2003, 2004; Younes et al., 2007).

Obstructive sleep apnoea is a common sleep disorder. A common feature in OSA is repeated hypopnoea or apnoea in the presence of breathing effort (Kapur et al., 2017), with concurrent oxygen desaturation (Safar et al., 1959). Clinically it can result in daytime hypersomnolence and has associated signs of disturbed sleep such as snoring and restlessness. The prevalence of OSA varies depending on the definition, but is estimated to be between 3 – 7 percent in the general population (e.g. Kim et al., 2004; Peppard et al., 2013; Punjabi, 2008; Young et al., 2009; Young et al., 1993; for review, see Senaratna et al., 2017). Risk factors for OSA include advancing age, male gender, obesity, and upper airway UA anatomical abnormalities, with obesity being the strongest risk factor (Peppard et al., 2000b). A progressive increase in the prevalence of OSA is reported in association with increases in body mass index (BMI) and associated markers such as neck circumference and waist-to-hip ratio (Peppard et al., 2013; Young et al., 2004). In the four-year longitudinal

Wisconsin Sleep Cohort study, a 10% increase in weight was associated with a sixfold increase in odds of developing OSA (Peppard et al., 2000a). OSA is also two to three times more common in males, though this difference reduces in post-menopausal women (Quintana-Gallego et al., 2004; Tufik et al., 2010; Young et al., 2004). Anatomical risk factors such as craniofacial abnormalities, size of the lateral pharyngeal walls, tongue and UA soft tissue volume may have genetic links to OSA (Ciscar et al., 2001; Li et al., 2000), with one study suggesting increased soft tissue in UA may be a heritable trait (Redline et al., 1995). Increased neck soft tissues surrounding the UA will likely increased transmural pressure and increase UA collapsibility (Schwab et al., 2003). Obesity also leads to increased adiposity in the tongue muscle and soft tissues surrounding the UA, with a strong positive correlation between total structural and functional tongue volume with body weight (Stone et al., 2016). Craniofacial structure is similarly important for maintaining UA patency, with one study reporting increased mandibular length in males was associated with reduced risk of developing OSA but not in females (Chi et al., 2011). There are also racial differences in the prevalence of OSA, with African Americans more likely to develop OSA compared to Caucasian peers at a younger age (Redline et al., 1997), and also has more severe OSA at a latter age (Ancoli-Israel et al., 1995), independent of weight. Despite lower rates of obesity, OSA prevalence in Asia is similar to United States, suggesting UA anatomical differences as likely explanations for this racial difference (e.g. Dempsey et al., 2010; Epstein et al., 2009; Li et al., 2000; for review, see Malhotra and White, 2002). Vascular factors may also play a role in increased UA collapsibility during sleep. Previous experiments have reported changes in UA resistance due to vasoconstriction and vasodilatation (Wasicko et al., 1990; Wasicko et al., 1991). During recumbent sleep, studies have reported a reduction in OSA AHI by reducing rostral volume (Elias et al., 2012; Kasai et al., 2014; Redolfi et al., 2011; Redolfi et al., 2009). This may be particularly important to patients with high volume states such as congestive heart failure, end stage renal or liver disease. Other identified risk factors for OSA include nasal congestion, smoking (Wetter et al., 1994), substances and medications such as alcohol, benzodiazepines and narcotics (Mason et al., 2015).

The pathogenesis of OSA is multifactorial, and is likely due to unfavorable interactions between UA anatomy and sleep-related changes in UA behaviors, causing instability in the balance of UA patency (e.g. Dempsey et al., 2010; Isono et al., 1997) (Fig 1.4, 1.5). Different models have been proposed to explain this loss of balance of upper airway patency. One proposed model (Eckert et al., 2013) identified four key pathophysiological causes:

susceptible UA anatomy favoring collapse (high Pcrit) (Gleadhill et al., 1991); impaired neuromuscular drive or response of UA dilators during sleep (e.g. Jordan et al., 2007; Loewen et al., 2011; Mezzanotte et al., 1992; Remmers et al., 1978); lowered respiratory arousal threshold during sleep (e.g. Eckert et al., 2011b; Younes, 2004, 2008; Younes et al., 2007); and a ventilatory control system with high loop gain meaning that it has increased sensitivity such that the response overshoots the physiological target (Wellman et al., 2004; Younes, 2008; Younes et al., 2001). An unfavourable balance between the negative intraluminal pressure that acts to collapse the UA, and positive pressure generated by the upper airway dilatory muscles to open it (Remmers et al., 1978; Shepherd et al., 2006) will likely lead to development of airflow limitation. In order to maintain pharyngeal patency within and between breaths, delicate moment to moment coordination of UA muscles activity and drive is required (for review, see Bilston and Gandevia, 2014).

Figure 1.4 has been removed due to copyright restrictions. Figures are adapted from (Isono et al., 1997).

Fig. 1.4. Schematic model illustrating interplay between upper airway patency, upper airway muscle activities and luminal pressure.

The balance of positive dilatatory (UA muscles) and negative collapsing forces (inspiratory luminal pressure) will determine patency of the pharyngeal airway. This is dependent also on the intrinsic mechanical properties and behaviour of the passive upper airway (i.e. pharyngeal anatomy). (Adapted from Isono et al., 1997)

Fig. 1.5. Illustration of current upper airway biomechanical concepts.

The 3 most commonly described upper airway biomechanical models include (A) Starling resistor - a collapsible pharyngeal airway situated between the non-collapsible larynx and nasopharynx where airflow is dependent on balance of positive dilatory forces and negative collapsing force. (B) Reduced pharyngeal cross-sectional area due to increased pharyngeal soft tissue enclosed within a “rigid” bony structure. Airway patency can be affected by crowding of the soft tissue within the confined bony structure. (C) Tongue as a muscular hydrostat - a constant-volume structure, with compression in one location resulting in expansion in another part of the tongue. (Adapted from Bilston and Gandevia, 2014)

Physiological and histological changes also occur in the UA of OSA patients affecting both sensory and motor properties including temperature and vibratory mechanical stimuli thresholds (e.g. Boyd et al., 2004; Kimoff et al., 2001; Larsson et al., 1992; Lindman and Stal, 2002; Nguyen et al., 2005). Furthermore, other reflex responses may be delayed or altered in OSA and snoring subjects, such as swallowing (Jaghagen et al., 2000; Teramoto et al., 1999), the response to negative pressure pulse during wakefulness where there is increased genioglossus EMG activity but decreased palatal muscle EMG activity (Berry et al., 2003; Mortimore and Douglas, 1997). Although some of these changes may be partially reversible after continuous positive airway pressure (CPAP) treatment (e.g. Dziewas et al., 2007; Kimoff et al., 2001). During wakefulness, a higher level of phasic and tonic UA muscle activity is also recorded in OSA subjects compared to healthy controls (e.g. Fogel et al., 2001; Jordan et al., 2005; Mezzanotte et al., 1992; Suratt et al., 1988), with markedly reduced compensatory neural drive compared with healthy subjects at the onset of sleep (for review, see Malhotra and White, 2002). In two separate studies, Saboisky et al. further recorded earlier discharge, increased level of firing and longer duration of motor unit action potentials in particular groups of genioglossus inspiratory motor units, but not others, during quiet breathing in OSA compared to control subjects (Saboisky, 2008; Saboisky et al., 2012), and also showed evidence of chronic partial denervation and remodeling of the genioglossus motor units, possibly as a result of a chronic neuropathic insult secondary to OSA. This

concept of peripheral neuropathy is further supported by EMG studies in other oropharyngeal muscles and hypoglossal nerve conduction studies suggestive of a hypoglossal mononeuropathy in OSA subjects (e.g. Ramchandren et al., 2010; Svanborg, 2005). However, it is unclear whether the neuropathic process is the cause of, or as a result of OSA. It also remains unknown as to whether prolonged snoring and associated local vibratory trauma (Friberg et al., 1998) or nocturnal hypoxaemia (Ludemann et al., 2001; Mayer et al., 1999) induces the neuropathic change.

Histological studies also recorded changes in the properties of the UA muscle fibre types in OSA patients (Boyd et al., 2004; Woodson et al., 1991). In the genioglossus of OSA patients, changes in the distribution of type II muscle fibres in association with reduced proportion of type I fibres was also recorded when compared to control subjects (e.g. Carrera et al., 1999; Series et al., 1996), suggesting higher UA muscle fatigability in OSA patients (for review, see Eckert et al., 2013). This was entirely corrected by CPAP treatment (Carrera et al., 1999). Other studies found abnormal UA muscle fibre types distribution and atrophied angulated fibres in palatopharyngeal muscle of OSA subjects (Edstrom et al., 1992), and increased inflammatory cell infiltrates in mucosal and muscular layers of soft palate and tonsillar pillars in OSA subjects (Boyd et al., 2004). One possible explanation for the changes in fibre type composition is a compensatory dynamic and adaptive response of the UA muscles to increase demand (“load”) in order to maintain UA patency during repetitive pharyngeal narrowing that occurs in OSA (e.g. Petrof et al., 1996; Pette and Staron, 1997; Smith et al., 2005), similar to skeletal muscles response after resistive load exercise (Bowers et al., 2004). This is supported by a study of 30 min daily oropharyngeal exercises over 3 months that resulted in a significant reduction in severity of OSA (Guimaraes et al., 2009). The combined effects of intermittent hypoxia, repetitive mechanical loading, and vibration caused by snoring, due to reduction in pharyngeal airflow could explain the histochemical and morphological changes recorded in UA muscles of OSA subjects (for review, see Saboisky et al., 2015).

Recent development of hypoglossal nerve stimulation as a potential OSA treatment provides further understanding of the neuromuscular control and biomechanics of the UA musculature (e.g. Mwenge et al., 2013; Schwartz et al., 2012; Strollo et al., 2014; Van de Heyning et al., 2012; for review, see Dedhia et al., 2015; Mwenge et al., 2015; Schwartz et al., 2014; Zaidi et al., 2013). In animal studies, short term electrical stimulation induces an increase in the more fatigue resistant MyHC muscle fibre phenotypes in the genioglossus

(e.g. Hartmann et al., 1999; Hartmann et al., 2001; Pae et al., 2007) and styloglossus muscles (Kinirons et al., 2003). Early pilot studies in humans demonstrated variable degree of pharyngeal dilatation or “closure” and tongue movement depending on site of stimulation (e.g. Schwartz et al., 1996) with subsequent activation of different groups of tongue retractors and protruders. The motor response produced by electrical stimulation will depend on the particular sets of motor axons that are activated. Further studies confirmed selective stimulation of tongue protruder (genioglossus) or retractor muscles (styloglossus and hyoglossus) opens or closes pharyngeal airway respectively, but coactivation of both (especially the horizontal genioglossus fibres) resulted in lingual and pharyngeal stiffness (Eisele et al., 1997; Oliven et al., 2007), providing multilevel UA improvement and stability of pharyngeal patency. Another recent development is the use of transcutaneous electrical stimulation of the upper airway to stimulate the dilatory muscles during sleep (e.g. Hida et al., 1994; Miki et al., 1989). Intermittent transcutaneous electrical stimulation was previously shown to reduce the AHI and improved oxygen saturation, but its clinical application was limited due to arousals when higher electrical stimulation frequencies were used (Edmonds et al., 1992). Two recent feasibility studies using continuous low electric current stimulation in OSA subjects reported reduction in oxygen desaturation index but not the AHI or snoring duration, and did not significantly affected the sleep quality (Pengo et al., 2016; Steier et al., 2011). Further studies are needed to determine likely responders in a clinical setting.

Recent studies in OSA subjects have revealed a dissociation between the EMG of tongue muscles and their mechanical responses during pharmacologically induced sleep (Dotan et al., 2015; Dotan et al., 2013). Compared to wakefulness, large increases in phasic genioglossus EMG triggered by prolonged airflow limitation during propofol-induced sleep failed to restore upper airway patency (Dotan et al., 2013). Furthermore, during a hypopnoea/apnoea event, EMG activities in the styloglossus and hyoglossus, two tongue retractor muscles, were minimal and below levels recorded during wakefulness (Dotan et al., 2015). These two studies reveal the importance of a coordinated activation of upper airway protruders and retractors in maintaining pharyngeal patency during wakefulness and sleep.

To further understand the relative contribution of reduced UA muscle activity / responsiveness to loss of UA patency during sleep, it is important to establish any mechanical consequence to UA muscles. In Chapter 4, we investigate the influence of ventilatory drive and inspiratory mechanics on movement of genioglossus, the largest UA dilator muscle, during quiet breathing under ultrasound imaging.

3.6 Impact of aging on the upper airway

Aging is an important independent risk factor for the development of OSA, although the exact underlying mechanism is unclear (e.g. Young et al., 2002). Previous studies had demonstrated age-related changes in upper airway collapsibility (e.g. Eikermann et al., 2007; Kirkness et al., 2008; Marcus et al., 1999). MRI studies reported a lengthening of the soft palate in women as well as increased size of the parapharyngeal fat pads in both elderly men and women (Malhotra et al., 2006), and acoustic reflection technique reported reduced upper airway calibre (Martin et al., 1997), all supporting of a more collapsible upper airway. However no significant difference in the upper airway luminal volume is recorded under MRI (Boutet et al., 2016; Malhotra et al., 2006). Another potential factor is the finding of a lower arousal threshold associated with aging. Previous studies recorded increased frequency of spontaneous arousals and reduced upper pharyngeal muscle negative pressure reflex response during wakefulness with aging (e.g. Redline et al., 2004; Malhotra et al., 2006; Erskine et al., 1993). However a more stable and less sensitive ventilatory control system is reported in the elderly (e.g. Edwards et al., 2014; Wellman et al., 2007). Further studies are needed to determine the effects of aging on upper airway mechanics during wakefulness and sleep.

4. IMAGING OF UPPER AIRWAY IN NORMAL AND OSA PATIENTS

4.1 Imaging modalities of the upper airway

In the past, EMG of various upper airway muscles and differential pressure measurements were used to evaluate the upper airway (e.g. Horner, 1996; Hudgel, 1986; Kuna et al., 1997; Mezzanotte et al., 1992; Schwartz et al., 1988; Suratt et al., 1984; Wiegand et al., 1990). These provided understanding to upper airway physiology, but they were limited in their inability to provide adequate assessment of the soft tissues surrounding the upper airway, or evaluate the mechanical action of the upper airway muscles. Given the need to understand the morphology and mechanical behaviour of the bony and soft tissue structures and their roles in the physiology and pathogenesis of upper airway closure, various imaging techniques were developed. These modalities include cephalometric radiography, endoscopic procedures such as fibre-optic nasopharyngoscopy, fluoroscopy, different types of computed tomography (CT) such as 3D or conventional, magnetic resonance imaging (MRI), acoustic reflection and optic coherence tomography (for review, see Kuo et al., 2011; Schwab, 1998; Stuck and Maurer, 2008; Togeiro et al., 2010). A review of these modalities is provided below.

4.1a Nasopharyngoscopy

This endoscopic technique allows real-time view of the luminal surface of the upper airway beginning from the nasal passage. It allows static and dynamic anatomical assessment without irradiation. It is well tolerated and can be performed in an outpatient setting, but is invasive (Borowiecki et al., 1978; Ryan and Love, 1996; Sher et al., 1985). The other limitation is inability of this modality to view surrounding bony structures and tissues. It has been used to examine physiological changes in patients with obstructive and central sleep apnoea (Badr et al., 1995; Morrell et al., 1998; Woodson and Wooten, 1994), the hypotonic airway (Isono et al., 1997; Launois et al., 1993), changes in airway lumen shape and calibre post weight loss (Schwartz et al., 1991), surgery such as laser-assisted uvulopalatoplasty (Ryan and Love, 2000) or uvulopalatopharyngoplasty (UPPP) (Isono et al., 2003), or with using mandibular advancement splints (MAS) (Isono et al., 1995; Johal et al., 2005). It also has been used in both asleep and awake OSA subjects to predict suitability for UPPP (e.g. Hsu et al., 2004; Isono et al., 1999; Launois et al., 1993; Morrell et al., 1998).

4.1b Cephalometry

Lateral cephalometry involved radiographs of the head and neck with specific emphasis on bone and soft tissue landmark. It is a simple and well-standardised technique, and commonly used in clinical practice given its relative simplicity, accessibility, low cost and minimal radiation (Fleetham, 1992). It has provided important craniofacial information on the differences between normal subjects, snorers and those with OSA (Hou et al., 2006; Lowe et al., 1986; Sforza et al., 2000); between Chinese and Caucasian OSA subjects (Liu et al., 2000); and between obese and non-obese subjects (Ferguson et al., 1995; Mayer et al., 1996; Nelson and Hans, 1997; Pae et al., 1997). Both upright and supine positional differences have also been measured (Ingman et al., 2004). Reported determinants of apnoea severity include narrow posterior airway space, an enlarged tongue and soft palate, a small retrognathic mandible and an inferiorly situated hyoid (e.g. Bacon et al., 1988; Bacon et al., 1990; deBerry-Borowiecki et al., 1988; Shepard et al., 1991). The limitation is the need for irradiation, its 2-dimensional focus on soft tissue and bony structures, and subject requirement to remain awake, which does not allow real time dynamic assessment, especially during sleep. Cephalometry results are also not easily comparable. Although cephalometry has been used to predict the postoperative results of UPPP or assess suitability of patients for maxillomandibular advancement surgery, the predictive value is questionable (Doghramji et al., 1995; Hochban et al., 1994; Stuck and Maurer, 2008). Currently 3D cephalometry is being investigated as an analysis tool in orthodontics and potentially may be applied to upper airway image analysis (e.g. Zamora et al., 2011; for review, see Pittayapat et al., 2014).

4.1c Acoustic Reflection

This technique calculates the cross-sectional area of the upper airway through analysis of the phase and amplitude of sound waves reflected from surrounding airway structures (Fredberg et al., 1980; Jackson et al., 1977). This non-invasive and radiation free technique requires subjects to hold a device in their mouth which is connected to the acoustic reflection apparatus. The measurement is calculated against a defined distance from the mouth (e.g. Bradley et al., 1986; Brown et al., 1986; Hoffstein and Fredberg, 1991; Hoffstein et al., 1984; Rivlin et al., 1984). It has a high correlation with roentgenographic area. Using this technique, OSA subjects have smaller cross sectional airway area compared to normal subjects (Brown et al., 1985; Hoffstein et al., 1984; Rivlin et al., 1984). Furthermore, the technique also revealed pharyngeal cross sectional area is smaller in the supine posture as

compared with upright position, with greater changes in OSA subjects (Brown et al., 1987; Fouke and Strohl, 1987). When compared with non OSA subjects, OSA subjects have greater changes in pharyngeal area with lung volume and applied pressure (Brown et al., 1985). The technique also demonstrated gender differences in airway dimension, with men having a larger upper airway in a seated position (Martin et al., 1997). However, this technique now has limitation clinically. The use of a mouth-piece makes it difficult to compare data to other imaging modalities where the mouth is closed. It also does not provide direct imaging of the upper airway, surrounding soft tissues, or craniofacial structures, limiting information only to airway lumen. Although a research device applying acoustic signal through the CPAP tubing to tract upper airway patency was recently developed (Stockx et al., 2010), better imaging techniques are now available.

4.1d Fluoroscopy

Fluoroscopy provides a dynamic evaluation of the upper airway during wakefulness and potentially during sleep. It has been used to study upper airway closure during sleep in patients with OSA. It demonstrated that in most patients, airway closure occurs in the retropalatal region but adequate cross-sectional measurements of the airway are not possible (e.g. Fleetham, 1992; Katsantonis and Walsh, 1986; Rojewski et al., 1982; Suratt et al., 1983). Currently its use in clinical setting is limited by the significant radiation exposure and its poor sensitivity to measure changes in airway size or provide detailed dynamic motion of soft tissue structures surrounding the upper airway.

4.1e Computed tomography (CT)

CT is an imaging technique performed in a supine position, and uses computer-processed x-rays to produce cross-sectional (tomographic) images of the targeted area. It is non- invasive, provides excellent imaging of the airway, soft tissue, and bony structures from the nasopharynx to larynx, and can measure upper airway cross-sectional area accurately. Using different CT scanners, besides axial images, three-dimensional reconstruction of the upper airway (Li et al., 2003a; Ryan et al., 1991b), dynamic evaluation (electron beam CT), and constructing volumetric images using helical CT scanners (Akan et al., 2004; Aksoz et al., 2004) can also be performed. Studies utilising CT have demonstrated narrowing of the upper airway in OSA subjects, especially in retro-palatal region (e.g. Burger et al., 1992a; Burger et al., 1992b; Ell et al., 1986; Ell et al., 1985; Haponik et al., 1983; Kuna et al., 1988; Ryan et al., 1991b; Shepard et al., 1990b). CT also is used to determine the width of the tongue and

upper airway muscles. Electron beam CT with dynamic imaging has also demonstrated changes in airway caliber during respiration in awake humans. The dimension is relatively constant during inspiration, enlarges in early expiration and becomes smallest at the end of expiration in healthy and OSA subjects (Galvin et al., 1989; Schwab et al., 1993; Shepard et al., 1990a). CT also has been used to identify suitable UPPP candidates, primarily those with retropalatal obstruction identified on CT rather than retroglossal obstruction (Ryan et al., 1991b; Shepard and Thawley, 1989). Cine CT or ultra-fast CT can obtain multiple images with a lower radiation exposure than standard CT (e.g. Shepard et al., 1990b; Stein et al., 1987; Suratt et al., 1983). However, limitation of CT included limited soft tissue contrast resolution when compared to MR imaging, its high cost, and the exposure of participants and patients to radiation.

4.1f Optical coherence tomography (OCT)

This is a light-based non-invasive imaging technique using broad-bandwidth light sources to provide high-resolution 2D or 3D cross-sectional images (Huang et al., 1991; for review of technique, see Podoleanu, 2005). The technique requires direction of an optical beam at an optical scattering media (e.g. biological tissues), and receiving any reflected light from sub- surface features, limited to a depth of one to two mm. It has been applied extensively in the field of ophthalmology to obtain high-resolution images of retina, in the field of cardiology to image coronary arteries and also in dermatology and oncology (for review, see Fercher et al., 2003; Tomlins and Wang, 2005).

In its application to upper airway imaging, Armstrong et al. placed a thin, transparent catheter through the nose into the upper airway. An adaptation to increase the depth of imaged distance to 26 mm was performed to allow better visualisation of pharyngeal anatomy (Armstrong et al., 2003; Armstrong et al., 2006). The method has high intra- and inter-rater reliability, with a reproducibility ICC greater than 0.7 in measuring cross-sectional area of velo-, oro- and hypopharynx and has excellent comparability with CT. In comparing awake OSA subjects and healthy subjects with OCT, velopharynx was identified as the narrowest upper airway site, and OSA subjects has smaller minimum cross-sectional area with increasing severity (Walsh et al., 2008). Currently OCT is used as a research tool to image upper and lower airway wall structure both ex vivo (Hariri et al., 2013a; Hariri et al., 2013b) and in vivo (e.g. Lam et al., 2008; Tsuboi et al., 2005), to provide measurement of cross- sectional airway size (e.g. Armstrong et al., 2006), and to assess upper and lower airway

compliance in different pathological states (e.g. Walsh et al., 2008; Williamson et al., 2010; for review, see McLaughlin et al., 2014). Although the technique can provide excellent resolution in real-time, the accuracy is limited by the superficial image penetration depth. Currently the lack of commercially available pulmonary system limits its usage outside of research setting.

4.1g Magnetic resonance imaging (MRI)

MRI is currently considered the best upper airway imaging technique as it provides excellent upper airway and soft tissue resolution (including adipose tissue), accurate quantification of cross-sectional area and volume, and ability to image in the axial, sagittal, and coronal planes with three dimensional volumetric reconstructions of soft tissue and bony structures (e.g. Abbey et al., 1989; Redline et al., 1992; Rodenstein et al., 1990; Ryan et al., 1991a; Shelton et al., 1993a; Shelton et al., 1993b). Although the associated noise and arousals during MR imaging can affect sleep onset, MRI has been performed during wakefulness and sleep without exposing subjects to radiation (Okada et al., 1996; Strobel and Rosen, 1996). Given its qualities, researchers have been able to use MRI to further understand pathogenesis of OSA and mechanisms of OSA treatments including UPPP, nasal CPAP therapy, weight loss and MAS (e.g. Rodenstein et al., 1990; Ryan et al., 1991a; Schwab et al., 1995; Schwab et al., 1996; Shelton et al., 1993a; Suto et al., 1993; Trudo et al., 1998; Welch et al., 2002).

Recently dynamic, spectroscopic and tagged MRI (tMRI) studies have been utilised to further knowledge about the pathophysiology, tissue characteristics and biomechanics of the upper airway in patients with OSA (for review, see Schwab, 1998). Using ultrafast MRI, dynamic imaging can be obtained (e.g. Ciscar et al., 2001; Donnelly et al., 2003; Ikeda et al., 2001; Yokoyama et al., 1996), while spectroscopic MRI allows for quantitation of fat and water in soft tissues. To further understand the anatomical structure and properties in the upper airway and surrounding pharyngeal soft tissues, MR tissue tracking (“tagging”) technique was developed (Prince and McVeigh, 1992), with further development of a three- dimensional (3D) modeling method using spatial modulation of magnetisation method (SPAMM). In this method, the MR scanner creates a grid of temporary “tags” (saturated magnetization) within the imaged tissue. This then allows determination of the grid deformation with tissue motion through ultrafast MR imaging.

Using tMRI and SPAMM, initial studies were aimed to investigate tongue movements in relation to speech (Napadow et al., 1999b; Niitsu et al., 1994; Stone et al., 2001). Subsequent studies have measured pharyngeal wall tissue movements in rats during bilateral medial hypoglossal nerve stimulation, demonstrating significant oropharyngeal and velopharyngeal dilatation with associated displacement and strain in discrete parts of the pharyngeal wall (Brennick et al., 2004); and comparing effect of obesity on upper airway size and pharyngeal wall tissue strain in rats (Brennick et al., 2006). Using the same technique, the mechanical behavior of a rat pharynx was demonstrated during applied positive and negative pressures to have the largest displacement occurring in the tongue and relatively small motion of the lateral pharyngeal wall tissues. This allowed for the development of a predictive model of upper airway mechanics (Xu et al., 2009). tMRI studies in human during quiet breathing in a supine posture have also demonstrated anterior movement in the genioglossus during inspiration with little movement in the surrounding upper airway tissues (Cheng et al., 2008) (Fig. 1.6). Furthermore, less motion was observed when inspiratory resistive load was applied suggesting an increase in the collapsibility of the lateral airway walls at the level of the soft palate (Cheng et al., 2011a). In awake OSA subjects, overall 4 movement patterns in the posterior tongue have been described during inspiration under tMRI. However, minimal inspiratory movement in the posterior tongue and lateral walls of nasopharynx was observed in severe OSA subjects with an apnoea/hypopnoea index (AHI) > 50, while those with more intermediate AHI demonstrated a more heterogeneous movement pattern, and normal subjects demonstrating more en bloc anterior motion (Brown et al., 2013a).

More recently, MR elastography (MRE) a new imaging technique that allows measurement of the viscoelastic properties of soft tissues in-vivo, has been applied to measurement of stiffness of the upper airway (Basford et al., 2002; Muthupillai et al., 1995). This technique has the potential to define a biomechanical model of the tongue and upper airway and elucidate mechanisms of different respiratory disorders (Buchaillard et al., 2007; Gerard et al., 2005; Stavness et al., 2008). MRE has been used to measure the viscoelastic properties of the tongue and soft palate in healthy volunteers (Cheng et al., 2011b) and OSA subjects (Brown et al., 2015) during quiet breathing, which may in future allows measurement of the active and passive properties of the tongue during respiration. Finally, recent development in 3D high resolution MRI technique further allowed calculation of functional and structural tongue volumes as well as anatomical differences within tongue muscles (Stone et al., 2016; Woo et al., 2015; Woo et al., 2012). A 4D MRI method

combining tagged and cine MRI images of the tongue has recently been developed to quantify the motion (Xing et al., 2017).

Fig. 1.6. Tracked movement of a grid of points on the tongue during a respiratory cycle in a single subject during a MR tagging study.

An example of a tagged MR image in the mid-sagittal plane (A) and lower axial plane(B) with a superimposed grid of points tracking the motion of the tongue during a respiratory cycle. Distinct respiratory movements (enlarged panels) are observed in the posterior region of genioglossus, where largest motion was recorded (≈ 1 mm antero-inferiorly during inspiration). Genioglossus was recorded to move anteriorly during inspiration and posteriorly during expiration. (Adapted from Cheng et al., 2008)

Although static and dynamic MRI has improved our knowledge of the pathophysiology of upper airway disorders and in particular OSA, MRI is currently used more as a research rather than a standard clinical tool due to the technique’s limitations. These include the difficulty in performing MRI during sleep (especially spontaneous sleep), limited availability of MRI, and the high associated costs, exclusion of patients suffering from claustrophobia or those with ferromagnetic clips or pacemakers. In Chapter 3, we report a comparative study between a tMRI technique with our novel ultrasound technique to quantitate the tongue motion in healthy and OSA subjects.

4.2 Upper airway imaging during sleep/wake states

Through the different modalities, assessment of the UA morphology has been performed in states of awake, asleep, and during respiration (for review, see Schwab, 2001; Stuck and Maurer, 2008). Predisposing static factors (craniofacial and soft tissue structures) to OSA have been examined during wakefulness, while studies during sleep provide information to state-dependent upper airway anatomical-physiological changes (for review, see Chen et al., 2016). Studies during respiration allows dynamic assessment of the UA to identify factors contributing to occlusion or collapse in OSA compared to normal subjects (for review, see Schwab and Goldberg, 1998).

4.2a Image studies during wakefulness

Compared to normal individuals, OSA subjects have smaller UA (e.g. Ciscar et al., 2001; Horner et al., 1989; Mezzanotte et al., 1992; Rivlin et al., 1984; Schwab, 1996). The retropalatal oropharynx has been demonstrated to be narrowest part of the UA in most patients (Galvin et al., 1989; Schwab et al., 1995; Shepard and Thawley, 1989), with the retroglossal region identified as the primary site in small minority of OSA subjects (Bhattacharyya et al., 2000). The main cause of this is a combination of a relative or actual excess of UA soft tissue in the space surrounding the pharyngeal lumen bounded by the bony structures, and / or altered bony framework itself (Schwab et al., 2003). Other contributing craniofacial features include shortened mandible length, inferiorly-positioned hyoid bone, and backward displacement of the maxilla (e.g. deBerry-Borowiecki et al., 1988; Lyberg et al., 1989a, b; Partinen et al., 1988; Pracharktam et al., 1994; Riley et al., 1983; Rivlin et al., 1984). It is also observed in two-thirds of OSA patients that the decreased pharyngeal size is associated with a shorter, more posteriorly positioned mandible (Riley et al., 1983; Shelton et al., 1993a). Other significant soft tissues abnormalities reported include increased volume in parapharyngeal fat pads, tongue, soft palate, and lateral pharyngeal walls with the likelihood of OSA correlating to the magnitude of volume increase in the latter three (e.g. Caballero et al., 1998; Ciscar et al., 2001; Do et al., 2000; Lowe et al., 1986; Ryan et al., 1991b; Schwab and Goldberg, 1998; Schwab et al., 1995; Schwab et al., 2003). However, the reported enlargement of lateral parapharyngeal fat pads does not always affect the airway lumen in OSA subjects (Schwab et al., 1995).

Further important differences in OSA subjects include a more anteroposterior orientation of the major pharyngeal airway axis compared to the lateral orientation in normal subjects. Primarily this is due to increased lateral pharyngeal wall thickness and may contribute to a compromise in maintaining airway patency by the pharyngeal dilator muscles (Leiter, 1996; Schwab et al., 2003). There is also a gender difference in normal pharyngeal airway length, soft palate cross-sectional area, and pharyngeal volume, with men having a larger and longer upper airway than women, with increased UA collapsibility at any given negative airway pressure (Malhotra et al., 2002; Martin et al., 1997). Craniofacial morphology differences were also observed across gender and race, and in first-degree relatives of non-obese OSA subjects compared to sex-, age-, weight- and height-matched controls (Mathur and Douglas, 1995; Will et al., 1995).

4.2b Imaging studies during sleep

Sleep onset is associated with a number of neurophysiological changes that affect neuromuscular activity, chemical, ventilation and load responses. When subjects transition from wakefulness to NREM sleep, there is a reduced central drive to the respiratory muscles and UA dilators, causing reduced UA cross-sectional area and increased UA resistance. This may compromise UA patency, but the location of obstruction during sleep is not necessarily the same as the location of UA narrowing during wakefulness (Hudgel, 1986; Suto et al., 1993). Imaging studies have shown that during sleep, the retropalatal region is the most common site of narrowing compared to retroglossal region, with airway narrowing occurring in the lateral and / or anteroposterior axes in OSA and normal subjects (e.g. Badr et al., 1995; Ciscar et al., 2001; Horner et al., 1989; Morrell et al., 1998; Suto et al., 1993; Trudo et al., 1998). In normal subjects, MRI studies demonstrate these changes are due to thickening of the lateral pharyngeal walls and posterior displacement of soft palate (Trudo et al., 1998). In OSA subjects, there is a combination of inspiratory and expiratory pharyngeal narrowing in the preceding breaths prior to an obstructive event. Furthermore, endoscopic and MRI studies have recorded increased UA collapsibility and reduced airway luminal size when compared to normal (Isono et al., 1997; Trudo et al., 1998).

4.2c Dynamic studies during respiration

Imaging studies have been used to study the UA during dynamic manoeuvres including inspiration, expiration, and Müller manoeuvre to simulate an apnoea (e.g. Burger et al., 1992a; Ciscar et al., 2001; Morrell et al., 1998; Ritter et al., 1999; Schwab et al., 1993;

Shepard et al., 1990b; Trudo et al., 1998). In these studies, the UA area remains relatively constant during inspiration due to balancing forces between airway dilators and negative airway lumen pressure. Maximal airway enlargement is observed in early expiration, most likely due to positive intraluminal pressure, while a reduction in UA luminal size is recorded at end-expiration. Other studies have also recorded apnoeic events during expiration, concluding that UA closures can happen throughout the breathing cycle (e.g. Badr et al., 1995; Morrell et al., 1998; Sanders and Kern, 1990; Sanders and Moore, 1983; Schwab et al., 1995). While airway narrowing is not homogeneous in normal subjects in studies employing Müller manoeuvre, OSA subjects do have greater narrowing when compared with normal (Ritter et al., 1999). Furthermore, higher AHI is associated with more retropalatal narrowing during a Müller manoeuvre (Hsu et al., 2004).

Recently under tMRI studies, anterior movement of genioglossus was observed during inspiration in normal subjects during quiet breathing (Cheng et al., 2008, 2011a). The location with most mechanical deformation appeared to be localised to a region just above the geniohyoid and anterior to the epiglottis. The magnitude is decreased during a loaded inspiration, together with a decreased dimension of the lateral airway. Genioglossus motion is also recorded to begin in early inspiration, most likely related to its action to dilate the UA and consistent with its EMG activity (e.g. Saboisky et al., 2006). In a similar MR tagging study, this inspiratory movement varied both within and between subjects with and without OSA (Brown et al., 2013a), with 4 types of posterior tongue movement patterns described, (i) “en bloc” anterior displacement of more than 1 mm, (ii) minimal displacement, (iii) oropharyngeal region displaced great than 1mm only, (iv) bi-directional movement with anterior displacement greater than 1mm in oropharyngeal region, and posterior displacement great than 1 mm in nasopharyngeal region. However, in OSA subjects with AHI > 50, the observed genioglossus movement was minimal. In Chapter 3 and 5, further description of posterior tongue movement during inspiration in healthy and OSA subjects under ultrasound imaging is reported.

5. ULTRASOUND IMAGING OF THE UPPER AIRWAY

Due to the non-invasive nature of ultrasonography, it has been increasingly used to image the upper airway both experimentally and in clinical setting by anaesthetists and critical care physicians. Comparing to other imaging modalities, ultrasound is shown to be safe, non- invasive and provides real time imaging.

5.1 Diagnostic ultrasound physics

Diagnostic ultrasound (US) is based on the properties of high frequency sound waves in tissues and does not emit electromagnetic radiation. The frequencies are between 1 – 20 MHz (million cycles per second), much higher than human audible range (20 – 20,000 Hz). A transducer converts electrical energy into pulses of ultrasound waves, and their propagation causes alternating compression and rarefaction of the conducting tissue. The velocity of the US beam is assumed to be 1,540 m/s in human tissue, which allows the calculation of the wavelength by the equation:

λ (wavelength) = c (speed) / f (frequency)

5.2 Interaction with tissues by ultrasound waves

US waves propagate in a straight line through a homogenous medium. However, in a heterogeneous medium or at the interface between 2 different tissues, the US pulse interacts in a number of ways (Fig. 1.7).

1) Reflection – when the incident US wave directly deflect back to source like a mirror. How much reflection occurs depend on the tissue’s acoustic impedance

Acoustic impedance = tissue density × propagation velocity

The direction of the reflected echo is determined by orientation of the boundary surface to the sound wave.

2) Refraction – when the incident US wave passes into a medium with a different acoustic impedance and become deflected. 3) Scatter – when US wave encountered an interface which is “rough” or where the structure is small and < 1 wavelength of the US in dimension. This results in the US waves radiating in all directions (scatterers).

4) Attenuation - where US energy is absorbed / converted to heat energy. This is dependent on US frequency and wavelength, and acoustic impedance mismatches between interfaces.

Fig. 1.7. Schematic of tissue interactions with ultrasound wave.

Different types of interaction ultrasound beam have with different organs and tissues. Diffraction refers to spreading out of ultrasound waves as it moves away from the sound source. (Adapted from Anderson, 2002)

Except for reflection, the other 3 interactions reduce the magnitude of the US wave. The greatest cause of attenuation in the body is absorption, with sound energy is absorbed by the tissue and degrade into heat, thus heat is the commonest physical effect ultrasound has on the human body (Ziskin, 1993).

Tissues have different acoustic properties (Table 1.1). This resistance to passage of ultrasound wave is termed acoustic impedance and the larger the difference of the acoustic impedance between two tissues, the stronger will be the returned sound waves (Ziskin, 1993). Hence whenever the sound beam reaches the interface between two tissues, some of the energy is reflected and the rest transmitted across. Reflected sound waves are then received by the transducer and produce electrical signals depending on the magnitude of the sound wave. This allows creation of the US image. Image formation also requires determination

of depth of the tissue interface / structure. This can be calculated by detecting how long it takes for the wave to return to the transducer, with the following equation:

Depth = velocity × time / 2

Medium / tissue Propagation velocity (m/s) Attenuation (dB/cm/MHz) Air 330 10.00 Bone 4080 5.00 Muscle (transverse) 1580 3.30 Muscle (parallel) 1.30 Kidney 1560 1.00 Liver 1550 0.94 Soft tissue (average) 1540 0.70 Fat 1450 0.63 Blood 1570 0.18 Water 1480 0.00 Table 1.1. Common values for soundwave propagation velocity and attenuation in different medium/tissue. (Adapted from Goss et al., 1978)

Finally, resolution (ability to distinguish spatially close objects) of the image varies directly with the frequency and inversely with wavelength. However, the ability of the US wave to be sufficiently transmitted back to transducer (tissue penetrance) is affected by attenuation (Table 1.2). This is directly related to wavelength and inversely to frequency. The choice of US transducer frequency in clinical setting is often guided by these two factors.

Frequency (MHz) Attenuation coefficient Penetration (cm) (dB/cm) 2.0 1.0 30 3.5 1.8 17 5.0 2.5 12 7.5 3.8 8 10.0 5.0 6 15.0 7.5 4 Table 1.2. Common values for attenuation coefficient and penetration depth.

(Adapted from Kremkau, 2016b)

5.3 Diagnostic ultrasound equipment

US transducers contain piezoelectric crystals which deform and vibrate when excited. These alternatively compress and expand the applied electrical energy, converting it to US wave. After brief transmission (1 – 6 μsec) the transducer then act to receive and convert the reflected waves back to electrical energy, thus producing an image (e.g. Bushberg and Seibert, 2002; Hedrick et al., 2004; McDicken, 1991; Middleton et al., 2004). This image texture (speckle) results from the coherent ultrasound pulses transmitted into the body interacting with the structures of the imaged tissue, producing a characteristic brightness- mode (B-mode) image of the tissue.

A-mode (amplitude mode) is the simplest form of image display. The information is displayed two-dimensionally with amplitude on y-axis and tissue depth on x-axis. Amplitude is proportional to the incident angle and acoustic impedance.

B-mode (brightness mode) is the use of high amplitude echoes which create a larger electronic voltage through greater crystal deformation.

M-mode (motion mode) detects rapidly moving structures by rapidly alternating a single crystal between transmission and receive modes with rapid updating (>1000 Hz). This gives a very high temporal resolution.

There are 2 main types of two dimensional (2-D) transducers. Linear-array transducers produce a rectangular image, with the US pulse produced perpendicularly to transducer face. Advantages of these are large superficial field of view and high resolution in near field. Curved-array transducers provide a wider field of view and have a convex shape. Their advantage is to provide better imaging of deeper structures.

5.4 Image artifacts

In diagnostic ultrasonography studies, sound beams will often transverse through different biologic tissues with different acoustic impedance. The subsequent reflection, refraction, scattering and attenuation at those interfaces give rise to different distinct imaging patterns.

Acoustic shadowing occurs as a result of reduced transmitted sound wave due to reflection and / or attenuation.

Ring-down artifacts are solid streak or series of parallel bands radiating away from a particular interface when there is a large acoustic impedance difference across the 2 interfaces (Avruch and Cooperberg, 1985), especially between air and water.

Another reverberation artifact is the comet tail artifact, which is usually associated with foreign bodies (Ziskin et al., 1982). These occur due to multiple strong reflections from the same surface. The spacing between the echoes gets narrower the smaller the object is, and more reverberant echoes occur where there is larger acoustic impedance difference.

5.5 Ultrasound image quality

Although a number of factors can affect the quality of the acquired ultrasound images, it can be summarised into two main categories.

1. Scanning approach of operator and how transducer is applied 2. Ultrasound machine settings and type of transducer chosen

5.5a Ultrasound machine settings

In B-mode ultrasound scanning, a number of machine settings can affect image qualities.

1. Gain - this control changes the amplification of all received echoes in the image, which affect the overall brightness of the image. A high gain will result in a bright (white) image, while a low gain will cause the image to turn dark (black). Inappropriate increase in overall gain can also cause US artifacts by increasing weak echoes from noise and boosting medium level echoes so they appear brighter, which reduce the contrast between strong echoes and those with lower amplitudes. 2. Depth – this setting alters the length of time the US machine waits for the returning echo. A deeper depth will reduce the frame rate and may affect image quality due to compromises in focusing, resolution, signal processing and line density. Depth and field of view should always be set appropriate to the tissue under examination. 3. Focus / focal zones – the focal zone is the particular depth or range of depths where spatial resolution is optimised. A higher number of zones can extend the range of depths, but comes at the expense of scan frame rate. 4. Time gain compensation (TGC) – due to attenuation of the US wave as it travels through the tissue, the US machine compensates for this by amplifying the more distant echoes to provide more uniform greyscales from similar tissue interfaces at

different depths. Most machines will allow an operator to manually modify the amount of amplification at different depths by adjusting a set of sliders or rotational controls representing depth of imaged tissue. Inappropriate usage of this function may adversely affect the intensity of the received echoes and create artifacts. 5. Digitising acquired US images – the US machine system or image analytical software may compress the images when digitising them. Compressed images (e.g. JPEG) are of poorer quality, but takes up less storage space. Uncompressed images (e.g. tiff or AVI movies), are of higher quality but may take up to 300 – 1200 KB/frame (Stone, 2005).

5.5b Operator and tissue dependent factors

1. Tissue properties – propagation of US wave is usually non-uniform within a tissue and can be enhanced or decreased. A tissue-tissue interface that is perpendicular to the US beam produced the best image quality due to complete reflection compared to one which is close to parallel to the US wave which causes refraction. An uneven tissue edge will also cause reflection in multiple directions and may reduce image quality. 2. Subjects – image quality often vary across different subjects. Increased fat content in the imaged tissue may cause increase refraction and thus poorer image quality. 3. Transducer placement – best image quality is achieved when transducer is positioned intimately with the skin and stabilised during scanning.

5.6 Ultrasound of the airway

Initial upper airway US imaging is of soft tissue of the neck, and in particular pre-tracheal structure and anterior tracheal wall. It also was described in speech literature to study laryngeal activity (e.g. Hamlet, 1980; Hamlet and Reid, 1972; Holmer and Rundqvist, 1975), pharyngeal wall displacements (e.g. Ryan and Hawkins, 1976; Skolnick et al., 1975), and tongue movements (e.g. Kelsey et al., 1969; Minifie, 1973; Sonies et al., 1981; Watkin and Zagzebski, 1973; for review, see Kelsey et al., 1969). In recent years, ultrasonography has also been used to provide rapid assessment of airway anatomy by anaesthetists and critical care physicians (e.g. Singh et al., 2010; Sustic, 2007; Tsui and Hui, 2008, 2009), and provided good comparison to CT in assessment of upper airway anatomic parameters (Prasad

et al., 2011). Ultrasonography also has been used to evaluate OSA (Siegel et al., 2000), and assessing patients for presence or absence of OSA (Lahav et al., 2009; Shu et al., 2013).

In upper airway US imaging, the transducer can be orientated in sagittal, parasagittal or transverse position (for review, see Kundra et al., 2011). The tongue tissue-air interface is an excellent reflector of US beams, with 99% of sound beams reflected back (Stone, 2005) (Table 1.3). The low acoustic impedance of air/gas results in scattering of sound waves and produces a dark appearance on the monitor. This results in poor imaging of the palate if the transducer is held submentally against the skin. Intraluminal air also results in difficulty to visualise deeper structures when imaging oral and nasal cavities, pharynx, larynx, and tracheal space. Under US imaging, bony structures will appear as bright hyperechoic linear structures with an underlying hypoechoic acoustic shadow due to reflection of sound waves, while cartilaginous structures are homogeneously hypoechoic. The hyoid and mandible bone-soft tissue interfaces thus cause significant refraction of the sound beam resulting in acoustic artifact, which may affect image quality. Muscle fascicles can be visualised as hypoechoic cylindrical structures, with surrounding connective tissue producing a hyperechoic appearance. Fat and glandular structures often appear homogeneous and appear more hyperechoic in comparison with adjacent soft tissues, depending on the fat content. Due to the large acoustic impedance difference, air–mucosa interface often produces a bright hyperechoic linear appearance.

Tissue interface Percentage of incident energy reflected

Fat/muscle 1.08

Muscle/blood 0.07

Bone/fat 48.91

Fat/kidney 0.64

Soft tissue/water 0.23

Soft tissue/air 99.90

Table 1.3. Percentage of incident energy reflected at tissue interfaces assuming flat interface and perpendicular incidence.

(Adapted from McDicken and Anderson, 2011).

5.7 Ultrasound imaging of the tongue

Ultrasound imaging of the tongue can be easily performed by applying the transducer on the submental skin in B-mode (Lahav et al., 2009). The transducer can be orientated in a sagittal or transverse direction, and the tongue is visualised deep to the muscles of the floor of the mouth. Tissue properties of the tongue may degrade image quality by affecting propagation of soundwaves (Stone, 2005). These include amount of fat content within tongue which cause scattering of soundwaves; interdigitate arrangement of intrinsic and extrinsic tongue muscles (Stal et al., 2003) which may increase soundwave refraction due to increased number of mini tissue-tissue interface; and amount of moisture which can affect the reflection properties of the irregular tongue surface.

In the transverse view, the dorsal tongue surface appears as a curvilinear hyperechoic pattern due to an air-mucosa interface. Intrinsic tongue muscles have a striated appearance, and for the extrinsic tongue muscles only the geniohyoid, genioglossus and hyoglossus can be visualised, with the mandibular ramus and hyoid bone obscuring the imaging of the other muscles. In sagittal scanning, mylohyoid and geniohyoid appear as linear hypoechoic bands between the hyoid and mandible, with the genioglossus and hyoglossus deep to geniohyoid. The hyoid and mandible can create acoustic shadows which can obscure parts of the tongue (Stone, 2005). In sagittal plane, genioglossus appears as a fan-like structure originating from the mandible towards the dorsal surface of the tongue. Due to the strong tissue-air interface at the tongue surface, the palate cannot be visualised.

During ultrasound imaging of the tongue, it is important to stabilise both the transducer as well as head position, while maintaining a close contact between the transducer and the skin under the jaw (Stone and Davis, 1995). This should be done without depressing the tissue which can cause deformation of the tongue. In using a mobile transducer, measurements of the tongue are made in reference to the jaw (Stone et al., 1988). Head stabilisation further reduces error due to head motion or drift in position over time.

Ultrasound has been used to assess tongue movement during swallowing or speech in adult (e.g. Blissett et al., 2007; Cheng et al., 2002; Chiang et al., 2003; Peng et al., 2003; Peng et al., 2007; Saigusa et al., 2006a; Stone, 2005), during feeding in infants (e.g. Bosma et al., 1990; Fucile et al., 2009; Geddes et al., 2008; Miller and Kang, 2007), as well as measuring the size of tongue surface area (Davidson, 2006; Lahav et al., 2009) and tongue

thickness (Capilouto et al., 2012), with comparison of this tongue base thickness between healthy and OSA subjects (Chien et al., 2017). In Chapter 2, I report a novel method of using ultrasound and image correlation to quantify posterior tongue movement during inspiration. In Chapter 3, I report the level of agreement and consistency between MR tagging and US in measuring tongue movement.

5.8 Image correlation technique and its application

Based on different tissues producing different image speckle patterns, analysis of the image can discriminate between different normal tissues and to detect abnormal conditions based on a multidimensional feature space (e.g. Insana et al., 1986; Smith et al., 1982; Wagner et al., 1986). An automatic tissue characterisation algorithm can be formed based on these intrinsic backscatter properties of the tissues imaged.

A variety of speckle tracking methods have been used to quantify the deformation and strain of tendons and ligaments in characterising the behaviour of these tissues. The length of the contractile portion is a much-studied important property as it provides a measure of neuromotor drive in relation to the applied load (e.g. Bolsterlee et al., 2017; Herbert et al., 2011). It also allows differentiation between active portion of the muscle and its tendon, and the biomechanical behaviour of skeletal muscle in vivo (e.g. Herbert et al., 2015). Previously both in vitro and in vivo animal preparations were used to provide continuous measurement of contractile length (Rack and Westbury, 1974). These techniques include sonomicrometry which provide spatial and temporal resolution measurements (e.g. Carrier et al., 1998; Gabaldon et al., 2004; Gillis and Biewener, 2002), real-time and cine-phase contrast magnetic resonance imaging which measures velocity of muscle contraction (Asakawa et al., 2003) and behaviour of muscle fibres during shortening (Pappas et al., 2002), x-ray scanning and tracking method (Amis et al., 1987) and goniometry model to estimate muscle contractile length (Bobbert et al., 1986; Hawkins and Hull, 1990). Using these techniques, measurements of muscle length during movements has been measured in human quadriceps (Rutherford and Jones, 1992), brachialis (Herbert and Gandevia, 1995), gastrocnemius (e.g. Herbert et al., 2002; Heroux et al., 2016; Narici et al., 1996), triceps (Kawakami et al., 1998; Maganaris et al., 1998), and tibialis anterior muscles (Herbert et al., 2002; Ito et al., 1998).

Recently, ultrasound has been used to measure changes in position of visually distinctive features between two or more frames to calculate relative static contractile changes

(Maganaris, 2001; Magnusson et al., 2001). An automated tracking algorithm was subsequently described, successfully tracking gastrocnemius aponeurosis and free tendon movements over several millimetres to an accuracy of ± 0.2 mm (Magnusson et al., 2003). Another similar tracking technique based on spatial cross correlation was also developed, tracking changes in length of segments of muscle tissue in the contractile portion of the gastrocnemius and soleus muscle (Loram et al., 2006). Under nonbiological conditions, this technique is sensitive and able to resolve a low-frequency movement down to 1.5 μm, and as low as 5 μm in biological conditions with a temporal resolution of 80 ms. During these image correlation analysis, a series of control “tracking points” were placed in different regions of the imaged tissue in the first frame of each image sequence. The software program then defined a region of interest of (e.g. 31 x 31 pixels) surrounding each control point and tracked the displacements of these regions of interest across the image sequences using cross- correlation routines (e.g. Herbert et al., 2011). The program does this by centering a window (e.g. 61 x 61 pixels) on the location of the tracking point in the preceding frame of the sequence to define the search field for the location of the region of interest in the next analysed frame.

One limitation of the technique is it requires the presence of the same feature of interest in similar form in the two analysed frames. A change in the angle of pennation during muscle contraction may alter the fascicles’ orientation, resulting in a change of tracked feature (Bolsterlee et al., 2016b). This was demonstrated in a study on 2D US measurement of human gastrocnemius muscle, where US transducer orientation deviation of 15 degrees from the true fascicle plane resulted in errors of fascicle length and fascicle angle up to 14% and 23%, respectively (Benard et al., 2009). In another gastrocnemius US study, an increase from 0.4 mm per degree of misalignment to 1.1 mm per degree of misalignment was recorded when the US transducer is tilted 20º away from image plane perpendicular to skin (Bolsterlee et al., 2016a).

OUTLINE OF SUBSEQUENT CHAPTERS

To further understand the pathogenesis of OSA and the interplay between the balance of positive dilatory and negative collapsing forces, it is important to establish any mechanical consequence to UA muscles associated with reduced respiratory mechanics during wakefulness and sleep in healthy and patients with obstructive sleep apnoea. The genioglossus is the largest UA dilator muscle, and has been extensively studied using multiunit EMG in both humans (Fogel et al., 2001) and animals (Mezzanotte et al., 1992), but only recently is its motion during respiration been described and measured with MR tagging (e.g. Cheng et al., 2008). To further understand the genioglossus and any consequence of its mechanical response during wakefulness and sleep, and in subjects with or without sleep disordered breathing, an ideal imaging modality that is inexpensive, noninvasive, free of radiation, dynamic and provides real-time imaging to a subject in a supine position during sleep is required. Currently MRI is the best UA imaging modality but it is expensive, noisy and unable to be used with patients who have contraindications such as claustrophobia. Therefore the overall aim of this thesis is to develop a reproducible and robust ultrasound technique to measure dynamic genioglossal movement in healthy and OSA subjects.

In Chapter 2, the main aim is to develop a novel ultrasound method to measure genioglossal movement during inspiration in real time. The results confirm the ability of ultrasound to provide high-resolution dynamic recording and allows for measurement of this movement. We also demonstrated good intrarater repeatability within the same testing session, as well as across three different sessions with the technique.

In Chapter 3, the main aim is to to evaluate the agreement between ultrasonography and MRI in measuring regional tongue displacement in healthy and OSA subjects. Our ultrasound technique is compared with the published MR tagging technique to assess their consistency and agreement in measuring tongue movement during inspiration in wakefulness. The studied population includes healthy controls, as well as patients with various severity of OSA and BMI. The results demonstrate good consistency and agreement between the two, with ultrasound recording more variability across the posterior region of the tongue.

In Chapter 4, we aimed to measure genioglossus movement in different physiological conditions when ventilatory drive is altered. We recorded increased genioglossus displacement during voluntary hyperpnoea; decreased genioglossus anterior displacement and

increased inferior displacement during inspiration against a resistive load, and no change in genioglossus displacement when breathing at different lung volumes.

Finally, in Chapter 5, we record and measure tongue displacement of healthy control subjects and OSA subjects of different severity during wakefulness. The aim of this study is to assess the dynamic motion of the tongue muscle with ultrasonography in healthy and OSA subjects during quiet breathing, and to investigate factors that may influence this motion in OSA subjects. The original MR tagging study reported 4 different patterns of tongue movement in subjects with and without OSA, and in particular minimal movement in patients with severe OSA and en bloc anterior movement in healthy controls (Brown et al., 2013a). Based on our ultrasound methods, we found that genioglossus movement varies within and between subjects with OSA, but we did not find a significant relationship between any specific pattern of movement and the severity of OSA.

EXPERIMENTS

CHAPTER 2

A novel ultrasound technique to measure genioglossus movement in vivo

SUMMARY

Upper airway muscles are important in maintaining airway patency. Visualization of their dynamic motion should allow measurement, comparison and further understanding of their roles in healthy subjects and those with upper airway disorders. Currently, there are few clinically feasible real-time imaging methods. Methods such as tagged magnetic resonance imaging have documented movement of genioglossus, the largest upper airway dilator. Inspiratory movement was largest in the posterior region of genioglossus. This study aimed to develop a novel ultrasound (US) method to measure GG movement in real time. We tested 20 healthy awake subjects (21-38 years) breathing quietly in the supine posture with the head in a neutral position. Ultrasound images were collected using a transducer positioned submentally. Image correlation analysis measured regional displacement of genioglossus within a grid of points in the mid-sagittal plane throughout the respiratory cycle. Typically, motion began before inspiratory flow in an antero-inferior direction and peaked in mid inspiration. Average peak displacements of the anterior, posterior, superior and inferior grid points were 0.62 ± 0.34 mm (mean ± SD); 0.84 ± 0.53 mm, 0.51 ± 0.30 mm, and 1.00 ± 0.70 mm respectively. Largest displacements occurred within the most infero-posterior part of the grid (1.16 ± 0.83 mm). This method had good intrarater repeatability within the same testing session as well as across sessions. We have devised a simple non-invasive ultrasound method which should be a useful tool to assess genioglossus movement in normal subjects and those with sleep disordered breathing.

1. INTRODUCTION The human tongue forms a major part of the upper airway and has important roles in physiological functions such as respiration, speech and swallowing. Abnormalities of upper airway function are common and result in disorders such as obstructive sleep apnoea (OSA), speech impairment, and dysphagia. Anatomically, the tongue is composed of extrinsic and intrinsic muscles, with genioglossus (GG) an extrinsic muscle being the largest dilator of the upper airway. Anatomical studies of canine GG led to the hypothesis that human GG acts as the main protruder and depressor of the tongue body through horizontal and oblique compartments. (Mu and Sanders, 2000). GG electromyographic activity (EMG) increases just before inspiratory airflow (Hudgel, 1993; Saboisky et al., 2006; Saboisky et al., 2007b; Strohl et al., 1980), continues during inspiration and may be sustained tonically during expiration. Recently, tagged magnetic resonance imaging (tMRI) has documented that GG moves anteriorly and inferiorly to dilate the upper airway during quiet breathing with minimal movement of tissues in the lateral pharyngeal walls (Cheng et al., 2008). This movement begins before inspiratory airflow (Cheng et al., 2011a).

While treatment of upper airway disorders requires better understanding of the biomechanics of the upper airway, diagnostic imaging of the oral cavity is not used routinely due to several factors. Endoscopic modalities such as fiber optic nasopharyngoscopy are unable to demonstrate how each tongue muscle contributes to overall movement of the tongue in quiet breathing (Borowiecki et al., 1978; Ryan and Love, 1996; Sher et al., 1985). Fluoroscopy involves exposure to radiation and is unable to characterise the internal deformation of soft tissues of the tongue (Partinen et al., 1988; Riley et al., 1985). Magnetic resonance imaging (MRI), and computed tomographic imaging (CT) have also been used to evaluate the upper airway muscles and lumen diameters in normal and subjects with sleep- disordered breathing (Brown et al., 2013b; Caballero et al., 1998; Ciscar et al., 2001; Schwab et al., 1993). However, these techniques are expensive, time-consuming and less likely to gain widespread clinical use. CT also involves exposure to radiation. Ultrasound imaging (US) offers potential advantages for clinical imaging of the upper airway (Rhys, 2011) and has previously been used to demonstrate dynamic upper airway movements during swallowing and speech assessments (Shawker et al., 1983; Shawker and Sonies, 1984). It is non-invasive, does not emit radiation, can resolve soft tissue structures, and can display cross-sectional anatomy and tissue motion in real time (Shawker et al., 1984). A recent review also demonstrated good reliability of 2D US measurements of human limb muscle

fascicle lengths and pennation across a range of experimental conditions, including muscle contraction and relaxation (Kwah et al., 2013). The need for enhanced upper airway imaging was highlighted by recent tMRI reports of subjects with OSA studied during quiet breathing (Brown et al., 2013a) and mandibular advancement (Bilston and Gandevia, 2014; Brown et al., 2013b). In subjects with a low apnoea-hypopnoea index (AHI), there was simple forward movement of the posterior genioglossus with inspiration, while in very severe OSA (AHI >50/hr) inspiratory movement was minimal. Other studies have also noted abnormal GG electromechanical responses in OSA patients both awake and asleep (e.g. Dotan et al., 2013; Eckert et al., 2013; Gleadhill et al., 1991; Mezzanotte et al., 1992; Saboisky et al., 2007b).

Therefore, the present study was designed to assess ultrasonography as a method to image the human genioglossus. We hypothesised that similar to the tMRI technique (i) US can be used to quantify movement of the GG during quiet breathing in awake healthy subjects, (ii) the maximal displacement occurs in the infero-posterior part of the genioglossus, (iii) the major displacement is in an anterior direction during inspiration, and (iv) US imaging provides reproducible measurements of genioglossus movement across different imaging sessions.

This chapter has been published (Kwan et al., 2014). All results here and in subsequent chapters take account of a calibration error which is also published (Kwan et al., 2015).

2. METHODS Twenty normal subjects (10 males and 10 females) volunteered for this study. None had a history of major respiratory or sleep disorder. Subject characteristics are given in Table 1. All participants completed the Epworth Sleepiness Scale (Johns, 1991) and Berlin Sleep Questionnaire (Netzer et al., 1999). Ethics approval was granted by the Human Research Ethics Committees of the Northern Sector of South East Sydney Area Health Service and University of New South Wales. The study was conducted according to the Declaration of Helsinki (2008) and informed written consent was obtained.

2.1 Experimental protocol The experimental setup is shown in Figure 2.1. Measurements were made with subjects in the supine posture. Since flexion or extension of the head alters upper airway size (Amis et al., 1999), the head position was standardised with the Frankfort plane (defined by the

inferior borders of the bony orbits and the upper margin of the auditory meatus) perpendicular to the horizontal bed surface. In addition, the anteroposterior position of the head relative to the lower cervical spine was standardised. The angle between horizontal plane and a line from lateral angle of eye to tragus was constrained to 77 - 82°; and the angle between horizontal plane and a line from tragus to spinous process of C7 vertebrae to 37 - 42°. Angle measurements were made with the aid of 2 goniometers, one aligned along the horizontal plane and the other along the measurement plane.

Figure 2.1. Schematic representation of ultrasound image, grid and numbering system, and subject positioning. A: Ultrasound image with red line outlining the tongue and the position of the grid. Genioglossus (GG), mucosa (M), tongue surface (S), geniohyoid (GH), myohyoid (MH). B: Numbering system used with the grid. C: Subject positioning. Angle θ was between the horizontal plane and a line from lateral angle of the eye to tragus (A) and was set at 77 - 82º, Angle φ was between horizontal plane in and a line from tragus to spinous process of C7 vertebrae and was set at 37 - 42º.

A digital photograph of the head position of each subject was also taken for further analysis. Padding for the head, neck and shoulders maintained the desired head and neck posture. Subjects were asked to relax, remain awake and to breathe through their nose throughout the study. Subjects were asked to position the tongue in its usual ‘resting’ position, usually with its tip on the incisors. In this position, it is likely that the force-

generating capacity of the tongue was high, and has been described as the genioglossus optimum length (Saboisky et al., 2013; Sha et al., 2000). The resting ventilation of each subject was monitored using calibrated respiratory inductance bands (Inductotrace, Ambulatory Monitoring Inc., Ardsley, New York, USA) to ensure that consistent breathing could be achieved during ultrasonography. Signals were digitised at 1 kHz using a CED1401 data acquisition system and Spike2 software (Cambridge Electronic Design, Cambridge, UK). Respiratory data were analysed off-line to determine the inspiratory time, tidal volume and respiratory rate. The onset of inspiration was taken from the signal of the abdominal inductance band (Cheng et al., 2008, 2011a).

2.2 Ultrasound scanning and analysis Ultrasound images were collected using a Philips iU22 system (Andover, USA) with a curved array C8-5 transducer, which has a probe frequency of 5 to 8 MHz. To visualise the tongue, the handheld transducer was positioned submentally, aligned in the mid-sagittal plane and pointed cranially. This provided a lateral view of the tongue body, submental musculature and mandible. Similar images of the tongue have been described in previous work on ultrasound imaging of the tongue during speech (Shawker et al., 1983; Shawker and Sonies, 1984; Shawker et al., 1984), measurement of GG depth (Brown et al., 2011; Eastwood et al., 2003) and clinical imaging of upper airway (Rhys, 2011). The abdominal movement tracing was also recorded on the ultrasound images. Real-time B-mode images were collected for at least 5 consecutive stable breaths at a frame rate of ~40 Hz (termed a “sequence”). Three sequences were captured in each session. During scanning, time gain compensation, depth and near gain were adjusted manually for best image quality. The validity of US measurements of muscle has been established when the US transducer is aligned perpendicular to skin surface or adjusted to optimise image quality (Kwah et al., 2013). In our study, minor adjustment of the transducer position was performed for each subject to ensure the best image quality of the GG. The depth of the image acquisition was set to 6 cm. A scan of the surrounding tissues was also performed prior to recording to determine the position of the superior surface of the tongue, the mandible and the posterior portion of the tongue.

Image sequences were analysed off-line using custom image correlation software developed in MATLAB (Mathworks, MA). This tracked the movement of designated markers on the image throughout the video sequence (Bonnefous and Pesque, 1986; Dilley et

al., 2001). For regions of suitable image quality, a rectangular grid containing 3 columns (4 mm apart) of 5 points (3 mm apart) was then placed over the tongue as infero-posteriorly as possible to include the inferior/posterior region of GG. This region was selected as previous studies have shown that this region is the site of most local motion (Cheng et al., 2008, 2011a) (Fig. 1.1A). Theoretically, motion may also be due to the actions of the intrinsic tongue muscles (i.e., verticalis, transversus and longitudinalis muscles) and/or to action of extrinsic tongue muscles (i.e., hyoglossus and styloglossus). Previous EMG studies in human subjects documented respiratory activity in GG (Hudgel, 1993; Saboisky et al., 2006; Saboisky et al., 2007b; Strohl et al., 1980) and hyoglossus (Mateika et al., 1999). Although previous studies in rodents indicate minimal intrinsic tongue muscle activity during eupnoea (Bailey and Fregosi, 2004), similar studies in human subjects have yet to be conducted. For this grid, the 3 columns were placed in an anterior to posterior direction, with the 5 points placed from most inferior to most superior position. For analysis, points 1-5 constituted the ‘anterior’ group, points 6-10 the ‘middle’ group, and points 11-15, the ‘posterior’ group. Points 4, 5, 9, 10, 14 and 15 made up the ‘superior’ group. Points 1, 2, 6, 7, 11 and 12 made up the ‘inferior’ group within the grid. The nomenclature for the grid points is given in Fig. 1.1B.

Any sequence in which swallowing or jaw motion occurred was excluded. The sequence of 5 tidal breaths with the clearest images was then selected. Regional displacements over three of the five breaths in each sequence were analysed and then the average was calculated. The resultant displacements were calculated using Pythagoras’ theorem.

2.3 Repeatability To assess repeatability of results within a measurement session, the same rater analysed each ultrasound image sequence containing the 5 breaths for all 20 subjects. The mean peak resultant displacement of the ‘infero-posterior’ region within the grid (mean of the displacement of points 11 and 12) was compared between the 3 selected breaths within the same sequence. This region was selected because a previous study using MRI imaging showed prominent displacement occurred in this region (Cheng et al., 2008) and in the present study this region also showed the largest displacement in quiet breathing (see Results).

To assess repeatability between different sessions due to variation in subject posture, probe positioning, and image analysis, two further studies were performed in a subset of 10 randomly chosen subjects (5 males and 5 females). One study was conducted in the same session, in which the subject stood up, walked around for 5 minutes and was then repositioned and rescanned. The other study was conducted at least 1 week later. The mean peak resultant displacement of the infero-posterior region was compared between the 3 imaging sessions.

Finally, to assess intra-rater reliability, the same rater analysed one randomly chosen file from each of the 10 subjects on three different days beginning with grid placement on the image sequence. The mean peak resultant displacement of the infero-posterior region was calculated and this was compared between the 3 analysis sessions.

2.4 Statistical analysis Means and standard deviations (SD) were used for descriptive purposes in the text and figures. This convention is used throughout the thesis. To assess gender differences in subject characteristics we used unpaired t-tests. To determine whether there were differences in the degree of motion for different regions of the grid of points (anterior, posterior, superior and inferior regions) a one-way analysis of variance with repeated measures was used. Sphericity was not assumed and the Greenhouse-Geisser correction was used. Bonferroni’s correction for multiple comparisons was applied. The correlation between the mean peak antero-inferior displacement of the posterior genioglossus during the 3 separate tests was determined with the intraclass correlation coefficient (ICC) using the average two-way mixed reliability model (Shrout and Fleiss, 1979). The internal consistency of the mean antero- inferior displacement of the ‘infero-posterior’ region between the 3 measured breaths within one analysis session, and the internal consistency between the mean peak antero-inferior displacement of the posterior genioglossus during the 3 analysis sessions of the intra-rater reliability sub-study, were determined by the Cronbach’s alpha coefficient (Landis and Koch, 1977). We also performed correlation analysis between the resultant displacements and tidal volume as well as between resultant displacement and ratio between tidal volume and inspiratory time (indirect measure of respiratory drive). This was performed using Pearson’s correlation analysis.

Statistical analyses were carried out using IBM SPSS version 20. Statistical significance was accepted at p<0.05. An ICC > 0.70 and Cronbach’s alpha > 0.70 were considered to indicate acceptable reliability (Nunnelly, 1978).

3. RESULTS Subject characteristics are shown in Table 2.1. There was no difference between the genders for age, BMI, Epworth sleepiness score and Berlin questionnaire, but neck circumference was smaller in females (p<0.01). Head position was standardised with a mean angle of 40 ± 2° between the horizontal plane and a line from lateral angle of eye to the tragus and 79 ± 2° between the horizontal plane and a line from the tragus to the spinous process of the C7 vertebrae. Male (n=10) Female (n=10)

Age (years) 26.0 ± 5.6, (21 – 38) 25.9 ± 5.6, (21 – 37)

BMI (kg/m2) 21.8 ± 2.5, (18.5 – 26.1) 22.1 ± 2.3, (18.4 – 25.9)

Neck circumference (cm) 37.2 ± 1.8, (35 – 40.6) 31.8 ± 1.5, (30 – 34.3)

Epworth sleepiness score 5.3 ± 3.1, (2 – 10) 6.3 ± 4.4, (0 – 12)

Berlin questionnaire Low (n = 10) Low (n = 10)

Table 2.1. Characteristics of 20 subjects. Data are expressed as mean ±SD with the range given in brackets. Body mass index (BMI).

3.1 Visualisation of oral anatomy With the transducer held in a midline sagittal position, the bulk of the tongue was clearly visualised (Fig. 2.1A). GG appeared as a fan-shaped muscle in the anterior portion of the tongue and two components could be identified. The superior component terminated in the tongue body postero-superiorly at which point it was difficult to distinguish from the other tongue muscles. The smaller inferior portion appeared to be directed posteriorly in a more horizontal band toward the hyoid bone.

The mylohyoid and geniohyoid muscles were less echogenic (darker) distinct bands lying underneath GG. The tongue surface could usually be seen cranially as an echogenic line which indicated the junction between the superior surface of the tongue and air in the oral cavity. In most subjects, an echo-lucent band immediately beneath and parallel to the tongue

surface can be seen which represented the mucosal layer (e.g. Shawker et al., 1984). The posterior portion of the tongue could be visualised if the transducer was directed more posteriorly towards the hyoid bone. It gave a high- amplitude echogenic signal while the GG muscle usually would not be seen. The hyoid also produced an acoustic shadow.

Figure 2.2 Displacement of the infero-posterior region of GG in a typical subject over 3 imaging sessions. Respiratory cycle was determined by signals recorded from the abdominal respiratory bands. A: displacement in the anterior-direction; B: displacement in the superior-direction; and C: resultant displacement. Open arrow near the origin shows that GG movement occurred prior to the onset of inspiration.

3.2 Displacement during quiet breathing Regarding grid placement for image analysis, point 1 of the 15 grid points was placed 18.7 ± 5 mm posterior (mean ± SD) and 9.1 ± 3 mm superior to the internal mental spine of the mandible (insertion of GG). The average distance between the grid columns was 3.6 ± 0.5 mm and the average distance between the grid rows was 2.6 ± 0.3 mm. The grid occupied around 6% of the imaged tongue area. The mean tidal volume across the twenty subjects

during one imaging session was 389 ± 133 mL with a mean respiratory rate of 15 ± 4 breaths/min. For the breaths from which images were measured, the respiratory cycle averaged 4.27 ± 1.26 seconds, with a mean inspiratory time of 1.95 ± 0.57 seconds. Figure 2.2 illustrates the anterior, superior and resultant displacement of the infero-posterior region in a typical subject. To show the reproducibility of the movement, data are shown for the 3 imaging sessions (two on one day, and the third one week later). An example of the recorded motion sequence of a subject’s genioglossus under US is provided in Supplementary video files 1 and 2 embedded into a Powerpoint file.

Anterior Middle Posterior Superior Inferior Infero- group group group group group posterior group

Resultant 0.62 ± 0.79 ± 0.84 ± 0.51 ± 1.00 ± 1.16 ± 0.34 0.52 0.53 0.30 0.70 0.83

Male 0.69 ± 0.84 ± 0.87 ± 0.56 ± 1.01 ± 1.13 ± resultant 0.38 0.61 0.63 0.28 0.84 1.00

Female 0.54 ± 0.74 ± 0.81 ± 0.46 ± 0.98 ± 1.20 ± resultant 0.29 0.43 0.44 0.32 0.57 0.67 p value 0.35 0.67 0.82 0.43 0.92 0.86

Table 2.2. Average displacement of defined sub-groups in the posterior genioglossus. Average regional displacement (mm) of each of the pre-determined regions for 20 subjects. Data are expressed as mean ± SD. P values for comparison of data from males and females.

The maximal displacement during the inspiratory phase of the respiratory cycle averaged 0.71 ± 0.24 mm along the transverse plane (positive being in the anterior direction) and -0.16 ± 0.12 mm along the coronal plane (positive being in the superior direction). The mean resultant peak displacement was 0.76 ± 0.25 mm. Table 2.2 summarises the results for the 6 sub-groups within the 15-point grid (Fig. 2.1A). There was a significant difference in the magnitude of the peak resultant movement between the different defined groups in the postero-inferior portion of the tongue (F1.374, 24.984 14.919, p < 0.001). Overall, the largest displacement occurred in the infero-posterior group within the grid (~1.2 mm; Fig. 2.3; see also Table 2.3 for data from each point on the grid). Formal post-hoc testing revealed that movement in the posterior group exceeded that in the anterior group (p = 0.017) and that

movement in the inferior group exceeded that in the superior group (p = 0.001). There was no difference in the magnitude of GG motion between male and female subjects (p > 0.05, unpaired t-test). Figure 2.4 depicts the mean resultant displacement during inspiration in the tracked region across all subjects.

Fig 2.3. Mean displacement of the infero-posterior group within posterior GG grid for 20 subjects over a typical breath. Mean anterior, inferior and resultant displacement of the infero-posterior group within the posterior GG grid for each subject. Average motion for points 11 and 12 is shown. Each line shows the mean of 3 separate breaths for each subject.

Grid points Mean anterior Mean inferior Resultant displacement (mm) displacement (mm) displacement (mm)

1 0.75 ± 0.58 0.34 ± 0.31 0.86 ± 0.60 2 0.69 ± 0.47 0.20 ± 0.30 0.75 ± 0.50 3 0.56 ± 0.33 0.04 ± 0.22 0.61 ± 0.31 4 0.43 ± 0.33 0.03 ± 0.22 0.49 ± 0.31 5 0.45 ± 0.34 0.03 ± 0.17 0.50 ± 0.32 6 0.99 ± 0.86 0.40 ± 0.32 1.10 ± 0.87 7 0.97 ± 0.77 0.24 ± 0.24 1.03 ± 0.77 8 0.75 ± 0.56 0.15 ± 0.23 0.80 ± 0.56 9 0.58 ± 0.40 0.08 ± 0.22 0.62 ± 0.39 10 0.42 ± 0.27 0.08 ± 0.16 0.45 ± 0.28 11 1.05 ± 0.85 0.27 ± 0.32 1.13 ± 0.85 12 1.15 ± 0.81 0.27 ± 0.26 1.20 ± 0.82 13 0.79 ± 0.58 0.16 ± 0.19 0.82 ± 0.58 14 0.59 ± 0.42 0.10 ± 0.17 0.62 ± 0.41 15 0.44 ± 0.33 0.04 ± 0.16 0.48 ± 0.31 Table 2.3. Average displacement of each grid points for 20 subjects. For the analysis, points 1 - 5 constitute the anterior group, points 6 - 10 the middle group, and points 11 - 15 the posterior group. (Fig. 1B) Points 4, 5, 9, 10, 14 and 15 make up the superior group. Points 1, 2, 6, 7, 11 and12 make up the inferior group. Data are expressed as mean ± SD.

Figure 2.4. Resultant displacement of all points on the grid for GG. The mean resultant displacement and direction of each of the 15 grid points for the 20 subjects. A schematic of the grid, its size and location is included. The actual scale refers to the displacement of each grid point. Four points are numbered (1, 5, 11, and 15).

3.3 Timing of inspiratory displacement Inspiratory displacements of the ‘infero-posterior’ group detected with US occurred on average 140 ms prior to the onset of inspiratory airflow as estimated from the respiratory inductance bands. One subject’s inductance band tracing on the ultrasound imaging was poor and the result was not analysable. In all but one subject, anterior GG motion began before inspiration. Maximal resultant displacement of the ‘infero-posterior’ group was reached relatively early in the inspiratory phase for all subjects at an average of 1.1 seconds after the beginning of inspiration (27.4% of the respiratory cycle). In all subjects, anterior displacement of the ‘infero-posterior’ group was larger than the inferior displacement.

Using Pearson’s correlation analysis, there was no correlation between maximal resultant movement of the infero-posterior group within the posterior GG grid and tidal volume (Pearson’s coefficient = 0.129, p = 0.294). Similarly, no correlation was found between maximal resultant movement of the same group with the ratio between tidal volume and inspiratory time (i.e. mean inspiratory flow; Pearson’s coefficient = -0.12, p = 0.308).

3.4 Repeatability of ultrasound measure Within the same testing session, the consistency of peak displacement for all 20 subjects across three breaths was good (Cronbach’s alpha = 0.81). The ICC for repeatability of average peak displacement for the 10 subjects across three breaths in three separate imaging sessions over 2 separate days was 0.85. The intra-rater reliability for repeated measurement of the displacement of the same sequence in 10 subjects for 3 separate image analysis sessions was excellent (Cronbach’s alpha 0.99).

4. DISCUSSION This study describes a novel method of using US to measure movement of the GG, the largest tongue muscle and upper airway dilator, during quiet breathing in wakefulness. It was visualised as a large fan-like muscle originating from the mandible and extending upward into the tongue body. As it contracted over the respiratory cycle, it moved the tongue forward and downward and would thus act to open the pharynx (Launois and Whitelaw, 1990; Smith et al., 2005). This action is consistent with its anatomical description (see Dickson and Maue, 1970) and movement observed during quiet breathing with tagged MRI (Cheng et al., 2008, 2011a; Doran and Baggett, 1972). The US method is also demonstrated to be reproducible over different imaging sessions on different days. As noted in Methods,

contraction of both extrinsic and intrinsic muscles, along with local mechanical effects, may contribute to the measured motion within genioglossus (for review, see Bilston and Gandevia, 2014).

4.1 Genioglossus displacement during quiet breathing As hypothesised, the US imaging analysis revealed that the infero-posterior region in the posterior GG was displaced antero-inferiorly by ~ 0.8 mm during inspiratory phase of the normal respiratory cycle, with greater displacement in the anterior than the inferior direction. The infero-posterior region in the GG moved more (in both anterior and inferior directions) than other parts of the GG. These findings extend those obtained in a small number of subjects by Cheng and colleagues (2008, 2011a) with tagged MRI. They reported variable magnitude in anterior movement of posterior GG during inspiration among the subjects (1.02 ± 0.54 mm), with both the size of the movement and its variability being larger than in the present study. The inferior part of GG moved largely in the anteroposterior direction consistent with the direction of the lower GG fibres, with minimal movement of surrounding tissues.

This study also demonstrated GG movement begins prior to inspiratory airflow, typically by ~150 ms, although there is potentially some imprecision in the indirect measure of the timing of onset of inspiratory airflow of the order of ~100 ms. The maximal displacement of GG occurred ~1 second into inspiration, consistent with the phasic motion previously described by Cheng and colleagues (Cheng et al., 2008). It is also consistent with studies in which non-invasive sublingual electrical stimulation of GG was demonstrated to maintain upper airway patency in awake humans (Schnall et al., 1995) and invasive electrical stimulation of GG increased inspiratory airway patency and reduced apnoea severity in patients with obstructive sleep apnoea during sleep (Oliven et al., 2003). Both multiunit (e.g. Eastwood et al., 2003; Malhotra et al., 2000b; Mathew et al., 1982; Sauerland and Harper, 1976) and single motor unit studies (e.g. Saboisky et al., 2006; Saboisky et al., 2007b; Tsuiki et al., 2000) show GG EMG has both inspiratory phasic and a sustained tonic activity. The inspiratory phase activity increases ~ 100 ms before inspiratory airflow (e.g. Hudgel, 1993; Pillar et al., 2001; Saboisky et al., 2006; Saboisky et al., 2007b; Strohl et al., 1980) and continues with recruitment and rate modulation of phasic motor units of GG throughout the first half of inspiration. The onset of GG activity precedes that of inspiratory pump muscles (Saboisky et al., 2006). There is also a correlation between the increase in GG EMG and

negative airway pressure (Pillar et al., 2001) and hence the observed inspiratory anterior movement of the GG will counteract the effects of negative inspiratory pressure at the level of the epiglottis.

4.2 Application of ultrasound method This simple non-invasive novel US method quantified the GG movement in human subjects during quiet breathing. GG was easily visualised through sub-mental position of the transducer. We propose this can be done in subjects who are awake or during sleep, and the method is repeatable across different sessions on same and different days. The results show that it is a promising method for studies of GG movement (e.g. comparing healthy subjects to patients with upper airway or sleep disorders; patients with mandibular devices) because it provides quantification of tongue motion. Furthermore, there is evidence from MRI that respiratory motion of GG is markedly impaired in disorders such as obstructive sleep apnoea (Brennick, 2013; Brown et al., 2013b). So far, no imaging method for the oral cavity has become routinely useful, probably due to radiation exposure, cost, or poor dynamic real-time correlation with quiet breathing. Our method combined with physiological measures should be broadly applicable to patients with upper airway and sleep disorders.

4.3 Limitations Due to the paucity of data in utilising MRI or ultrasound to measure the magnitude of tongue movement in both healthy control and OSA subjects , a sample size of 20 subjects were chosen to participate in this study. The results can now be added to data for prospective sample size calculations in future studies.

As with all imaging techniques, US imaging of GG has limitations. Although resolution of the ultrasound image is excellent for air-tissue interfaces (e.g. upper and lower tongue surfaces), it is often difficult to visualise detailed intramuscular architecture of GG. Hence, improved linear probes and 3D reconstructions may prove beneficial. Such technical developments may make it easier to position the 15-point grid in a more precise manner, in relation to a bony reference point (e.g. mandibular plane).

Images obtained by US are sensitive to movements of both the subject and the transducer. Here, as the quantification of GG displacement relies on tracking of grid points throughout the respiratory cycle, small movements during collection of image sequences can degrade

image quality and affect the reproducibility of the imaging method. Therefore, standardizing the head position before US imaging and maintaining it during image capture is important. However, despite its limitations, reproducible measures of maximal inspiratory movement of GG were obtained within and across sessions, as indicated by high intra-session and inter- session reliability analysis in this study.

Because the imaging data was gathered during wakefulness it is unclear how applicable it is for sleep. Based on EMG measures, EMG activity of GG decreases during sleep onset, and thus GG movement may also differ (Wilkinson et al., 2008). The measured genioglossus motion shows that there can be apparent “drift” such that displacements do not always come back to zero at the end of each respiratory cycle. This could be due to breath-to-breath variability and/or errors by the program tracking points from frame to frame. Although our subjects breathed at a slightly faster respiratory rate and lower tidal volume, minute ventilation was maintained within population mean, and our results comparable to previous findings under tMRI. Finally, image correlation analysis has been validated in other skeletal muscles, for instance good accurate results were demonstrated for small displacements (1.0- 2.0 mm) in porcine muscle (Chen et al., 1995). The current method has not been validated in the human genioglossus muscle, where it is also difficult to define a standard bony landmark under US imaging in a sagittal plane, however the results are broadly consistent with the findings of Cheng and colleagues (Cheng et al., 2008). Furthermore, US storage systems commonly store DICOM images at a default setting of 30 frames/sec, which may underestimate deformation values (Koopman et al., 2010; Marwick, 2010) and affect image correlation analysis.

5. CONCLUSION In summary, this study demonstrated that US imaging is a repeatable and comparable method to existing imaging technique in measuring genioglossus movement. Compared to existing imaging techniques, US does not emit radiation and is quick to perform. It can offer real time imaging in both clinical and research setting. Its relative simplicity could provide a novel imaging technique for anatomical phenotyping in patients with upper airway disorder such as obstructive sleep apnoea.

CHAPTER 3

Sagittal measurement of tongue movement during respiration: comparison between ultrasonography and magnetic resonance imaging.

SUMMARY

The upper airway muscles are important for complex motor tasks such as swallowing, speech and respiration. The tongue is crucial in providing support and patency to the pharynx. MRI is currently considered the best imaging technique to evaluate upper airway soft tissue structures in disorders such as obstructive sleep apnoea. In healthy subjects during quiet inspiration, an anterior motion of 0.5 – 2 mm was recorded under tagged MRI and ~1 mm under ultrasound in the infero-posterior region of the tongue. This fits with the action of the genioglossus as the largest dilator muscle in the tongue. To our knowledge, this current study is the first to evaluate the agreement between ultrasonography and MRI in measuring regional tongue displacement in healthy and OSA subjects. In 21 subjects with and without OSA, we found good consistency and agreement between our ultrasound and MRI technique, with an overall intraclass correlation coefficient of 0.79 (95% CI: 0.75 to 0.82) between MRI and ultrasound in measuring anterior and posterior tongue motion during inspiration. Ultrasound revealed greater anterior displacement in the posterior tongue with a mean difference of 0.24 ± 0.64 mm (95% limits of agreement: 1.03 to -1.49). This may reflect the higher spatial and temporal resolution of the ultrasound technique. This study further validates ultrasound as a method to quantify inspiratory movement of tongue in healthy and OSA subjects, and confirms that maximal inspiratory displacement is in the infero-posterior region of the tongue.

1. INTRODUCTION The upper airway muscles are important for numerous complex motor tasks including swallowing, speech, mastication and respiration. During these motor tasks, the tongue provides support and patency to the pharyngeal space. Although the human tongue can move in multiple directions the major movements related to respiration are protrusion and retraction. Protrusion assists in dilation of the pharyngeal airway while retraction occludes it. In humans, the genioglossus muscle is the largest extrinsic muscle of the tongue, and the largest upper airway dilator (Mu and Sanders, 2000). It is active in speech, swallowing and respiration; and pulls the tongue antero-inferiorly during an isolated contraction to open the pharynx (Smith et al., 2005).

To understand the pathogenesis of upper airway disorders, various imaging techniques have been developed to assess the morphology and mechanical behavior of the upper airway bony and soft tissue structures. They include cephalometric x-ray, endoscopic procedures such as fiber-optic nasopharyngoscopy, fluoroscopy, different types of computed tomography, magnetic resonance imaging (MRI), acoustic reflection, and optical coherence tomography (e.g. Schwab, 1998; Schwab and Goldberg, 1998; Stuck and Maurer, 2008). Of these techniques, MRI is currently considered the best imaging technique for upper airway disorders such as obstructive sleep apnoea (OSA) (Brennick, 2013) as it provides excellent upper airway and soft tissue resolution (including adipose tissue), accurate quantification of cross-sectional area and volume, and the ability to image in the axial, sagittal, and coronal planes with three dimensional (3D) volumetric reconstructions of soft tissue and bony structures. It can be performed in wakefulness and sleep with no radiation exposure, and can track tissue motion.

Tagged MRI (tMRI) studies using spatial modulation of magnetisation method (SPAMM) (Axel and Dougherty, 1989; Axel et al., 1992) further improved our understanding of the properties in the upper airway and surrounding pharyngeal soft tissues. The posterior tongue was recorded to move anteriorly about 0.5 - 2 mm during inspiration in healthy awake subjects, with little movement of the surrounding upper airway tissues (Cheng et al., 2008). Maximal displacement was observed in the infero-posterior region. In awake OSA subjects, minimal movement during inspiration occurred in the posterior tongue and lateral walls of the nasopharynx in severe OSA subjects with an apnoea/hypopnoea index (AHI) > 50, while

those with intermediate AHI had a more heterogeneous movement pattern, and normal subjects had a more uniformed anterior motion (Brown et al., 2013a).

Although static and dynamic MRI can provide useful information about upper airway anatomy and function in disorders such as OSA, MRI is typically a research rather than a standard clinical tool due to the limited availability of MRI, high cost, exclusion of patients with claustrophobia or those with ferromagnetic implants or pacemakers. Compared to MRI and other imaging modalities, ultrasonography is safe, non-invasive and provides real time imaging in wakefulness and sleep. It is increasingly used to image the upper airway (e.g. Kundra et al., 2011; for review, see Stone, 2005), to evaluate tongue movement in adults (e.g. Cheng et al., 2002; Chiang et al., 2003; Peng et al., 2003; Peng et al., 2007; Saigusa et al., 2006a), to measure the surface area of the tongue (Davidson, 2003; Lahav et al., 2009) and tongue thickness (Capilouto et al., 2012; Chen et al., 2014).

In adults, the pathophysiological causes of OSA are likely multifactorial (Eckert et al., 2013), and it is crucial to understand each potential contributing factor including upper airway anatomy and collapsibility, and dilator muscle action. To date, ultrasonography has been used to identify and evaluate patients with OSA (Shu et al., 2013; Siegel et al., 2000). However, prior to tMRI and 2D ultrasound image correlation analysis, detailed image recordings and analysis of breath-to-breath tongue movement were not available. In Chapter 2, we reported a novel image correlation technique to measure regional displacement of tongue in vivo in the sagittal plane using an automated 2D tracking algorithm (Kwan et al., 2014; Kwan et al., 2015). During inspiration, the posterior tongue was observed to move antero-inferiorly by about 1 mm, with the largest displacement in the infero-posterior region. However, no studies have compared ultrasound and MRI methods for quantifying upper airway tissue motion to date.

Therefore, the aim of this study was to evaluate the agreement between ultrasonography and MRI in the measurement of regional tongue displacement on a sagittal slice in healthy and subjects with OSA during awake quiet breathing.

2. METHODS Twenty-one subjects (15 males and 6 females) with and without OSA volunteered for the study. None had a history of major respiratory disorder. The subjects were aged 20–68 years

and had body mass indices (BMI) ranging from 20.6 to 38.4 kg/m2. All participants completed the Epworth Sleepiness Scale (Johns, 1991) and underwent a diagnostic polysomnogram to confirm presence or absence of OSA. Diagnosis of OSA and determination of OSA severity were based on American Academy of Sleep Medicine guidelines (Berry et al., 2012). Ethics approval was granted by the Human Research Ethics Committees of the South East Sydney Local Health District and University of New South Wales. The study was conducted according to the Declaration of Helsinki (2008) and informed written consent was obtained.

2.1 Experimental protocol The experimental setup for both image modalities is shown in Fig. 3.1. Subjects lay supine and remained awake with their head in a standardised position with the Frankfort plane (defined by the inferior borders of the bony orbits and the upper margin of the auditory meatus) perpendicular to scanning table throughout both experiments. To minimise effects on upper airway size (Amis et al., 1999; Trudo et al., 1998) and genioglossus motion (Cai et al., 2016) due to head flexion or extension, head, shoulder and neck padding was used to maintain posture. Subjects were instructed to rest the tongue in its “normal resting’ position, usually with its tip on the incisors, and breathe through the nose. Frequent verbal communication with subjects occurred during the study to ensure wakefulness.

2.2 Ultrasound scanning and analysis Our ultrasound protocol (see Chapter 2) used a Philips Healthcare iU22 system (Andover, USA) with a curved array C8-5 transducer (frequency 5 to 8 MHz, aperture of 22.4 mm, 122º field of view) which provides good image quality and reliability to image the tongue movement during tidal breathing. To summarise, the transducer was handheld in the mid- sagittal plane, pointed cranially and positioned submentally. This allowed a lateral cross sectional view of the tongue body, submental musculature and mandible with the genioglossus appeared as a fan-like muscle originating from the mandible. Adjustments were made to time gain compensation, focus and near gain to optimise image quality, with the acquisition image depth set at 6 cm. Minor adjustment of the transducer position was made to visualise the mandible, the superior surface and the posterior portion of the tongue. The transducer was kept at a constant position and pressure with reference to the jaw during image acquisition. B-mode images were collected at a frame rate of ~40 Hz in real time. Resting ventilation of each subject was monitored simultaneously using thoracic and

abdominal respiratory inductance bands (Inductotrace, Ambulatory Monitoring Inc., Ardsley, New York, USA), with acquired signals digitised at 1kHz using a CED1401 data acquisition system and Spike2 software (Cambridge Electronic Design, Cambridge, UK). During the session, the abdominal inductance band output was displayed on screen to allow real-time recording of abdominal movement and determination of onset of inspiration. Each image “sequence” collected at least 5 consecutive stable consistent breaths, with each image collection session capturing three sequences. For analysis, the sequence of 5 tidal breaths with the clearest images was selected after exclusion of any sequence in which swallowing or jaw motion occurred.

Fig 3.1. Set up for ultrasonography and MRI studies including grid positions. Subjects lay supine in both experiments with standardised head position, with the Frankfort plane perpendicular to the scanning table. Calibrated respiratory inductance bands used to monitor resting ventilation during ultrasound experiment. Ultrasound transducer was positioned in the submental mid-sagittal plane. MRI and ultrasound images are of the same subject. Top left panel: MRI image of the placement of anterior and posterior grids. 15 solid black circles denoting the “tracking points”. 5 solid blue dots denote the “tracking points” at start of inspiration for posterior column. Top middle panel: Ultrasound image of the placement of the anterior and posterior grids with black line outlining the tongue and 15 solid black circles denoting the “tracking points”. 5 solid blue dots denote the “tracking points” at start of inspiration for posterior column. Top right panel: 15 grid points and columns represented. Points 1-5 defined as “anterior” group, 6-10 as “middle” group, 11-15 as “posterior” group. Points 11,12 defined as “infero- posterior” group. GG, genioglossus; GH, geniohyoid; MH, mylohyoid

Offline analysis of image sequences was performed by a single investigator using a custom 2D image correlation software developed in MATLAB (Mathworks, MA, USA). This technique relies on the principle that different tissues produce different image speckle patterns (Insana et al., 1986; Smith et al., 1982). During analysis, a series of control “tracking points” were placed in different regions of the tongue by using a cursor (see below section for location of tracked regions) in the first frame of each image sequence. The software program then defined a region of interest of 31 x 31 pixels surrounding each control point and tracked the displacements of these regions of interest across the image sequences using cross-correlation routines (Herbert et al., 2011). A window of 61 x 61 pixels centred on the location of the tracking point in the preceding frame of the sequence defined the search field for the location of the region of interest in the analysed frame. A 2D spatial filter was applied to smooth out the motion and to reduce random error (Garcia, 2010). Visual inspection of the tracking was performed and analysis was aborted if the tracked points appear to drift. Tracking points were then repositioned and reanalysed from beginning of the image sequence until investigator was satisfied accurate tracking of the image sequence was performed.

2.3 MRI scanning and analysis A 3T MRI scanner (Philips Achieva TX, Best, The Netherlands) was used to image tongue motion using a 16-channel transmit–receive neurovascular coil, with patient head positioned similarly to that used for ultrasound imaging. Respiration was monitored using a MRI- compatible respiratory sensor (Philips Healthcare, Best, The Netherlands).

Tongue motion in the mid-sagittal plane was imaged using a previously described 2D complementary spatial modulation of magnetisation (CSPAMM) imaging sequence (Brown et al., 2013a; Cheng et al., 2014; Cheng et al., 2008, 2011a). In brief, this technique involves “tagging” a rectilinear grid on the tissue by saturating the applied radiofrequency in a spatially selective manner around 140 ms before image acquisition. Images of the tagged tissue were collected every 250 ms until the grid faded from the region of interest (∼1000 ms after the tag application), then repeating this process to image over the respiratory cycle. The imaging parameters used were: flip angle 7 deg; repetition time 2.2 ms; echo time 0.9 ms; total scan duration 31 s; 256 × 256 matrix with a slice thickness of 10 mm (pixel size 0.86 × 0.86 mm2), and tag spacing 8.6 mm. Tissue motion was imaged continuously over 31 s (on average four times for each subject). Furthermore, a multi-slice turbo spin echo technique

(matrix 256 × 256, slice thickness 3 mm, 25 continuous slices, plane resolution 0.86 x 0.86 mm2, TR = 3000 ms, TE = 52 ms) was used to collect sagittal T1-weighted anatomical images of the upper airway. The total scan time was 108 s. In addition, axial T2-weighted anatomical images of the upper airway were also acquired (matrix 384 x 384, slice thickness 3 mm, 50 continuous slices, in-plane resolution 0.5 x 0.5 mm2, TR = 4965 ms, TE = 80 ms), total scan time 4 min 28 s.

Tagged image sequences with mandibular movement more than 1 mm were discarded. The harmonic phase method (HARP) was used to calculate tongue displacement (Osman et al., 1999). HARP uses the principle that tag spacing in the image is altered by tissue deformation, and this is quantifiable from changes in the spatial frequency of the tagged grid in the Fourier domain. It gives sub-pixel resolution with a displacement error of 0.1 pixels (Kerwin and Prince, 2000).

Fig 3.2. Sites of major anatomical measurement on MRI images. Left panel: sagittal MRI image shows pharyngeal length and tongue base angle. Thick blue line represented location used to measure pharyngeal cross-sectional area in axial view. Right panel: axial MRI image showing the pharyngeal cross-sectional area at level of soft palate, through the narrowest point of the nasopharynx.

Average anterior tongue motion during inspiration (beginning just prior to inspiration) was calculated over three respiratory cycles. In addition, the cross-sectional area of the pharynx at the level of the soft palate, pharyngeal length (vertical distance between the hard palate and the tip of the epiglottis), and the tongue base angle (defined by the angle between the inferior margin of the geniohyoid and the posterior pharyngeal wall) measurements were made from the anatomical images (Fig. 3.2).

2.4 Comparing regional movement during inspiration between MRI and ultrasound In the chosen MRI and ultrasound image sequences, tongue motion was analysed for each subject with two rectangular grids, each with 15 control “tracking points”. The grid was comprised of 3 columns (≈ 3 mm apart) of 5 points (≈ 2 mm apart). The nomenclature for the grid points is given in Fig. 1. One grid was placed in the posterior tongue and one in the anterior tongue in comparable regions for the two imaging modalities. These regions were matched by placing the bottom left corner of the grid at the same rostral and posterior distance from the internal mental spine of the mandible. The posterior tongue region is included as there is maximal inspiratory displacement with both MRI and US imaging, with non-uniform motion within this region (Cheng et al., 2008; Kwan et al., 2014). Based on prior studies, we expected minimal grid displacement in the anterior region during quiet breathing.

In the previous MRI study of OSA subjects, four patterns of sagittal tongue movement during inspiration were observed (Brown et al., 2013a). To compare the present data with this study, a column containing 5 additional “tracking points” (≈ 5 mm apart) was also placed in the most posterior region of the tongue for each subject in both modalities. Regional resultant displacements over three breaths in the chosen sequence of both techniques were analysed. To determine variability, agreement and consistency between MRI and US, the mean of the resultant displacements for each analysed region over 3 breaths was calculated for the 21 subjects.

2.5 Statistical analysis Descriptive statistics are presented as mean ± standard deviation. Categorical data were expressed as numbers and percentages and continuous data as medians and 1st – 3rd quartiles. Reproducibility of the ultrasound and MRI measurements of grid points was assessed using Bland and Altman analysis and a two-way mixed, absolute agreement, average measures intraclass correlation coefficient (ICC) (Bland and Altman, 1986; McGraw and Wong, 1996; Shrout and Fleiss, 1979). The median and interquartile range were calculated to assess variability between MRI and US results. To determine the sample size required to achieve a high level of agreement between measurements obtained from MRI and ultrasound, a pre- specified high ICC value of R0 at 0.90 in the null hypothesis (minimum level of agreement) and ICC value of R1 at 0.97 in the alternative hypothesis (targeted level of agreement) was

set. Therefore, based on only two observations made on each subject, a sample size of at least 18 is required to achieve a statistical significance for  = 0.05 and with a power of more than 80% (Bujang and Adnan, 2016). To assess sample size adequacy, a paired- samples t-test was conducted with calculation of 95% confidence interval of the mean difference between MRI and US results for posterior tongue grid. We further used the coefficient of variation to determine variance between MRI and US results, and between each of the 3 breaths measured using the US technique across the 21 subjects. Statistical analyses were carried out using IBM SPSS Statistics for Windows, Version 24.0 (Armonk, NY, USA); MedCalc for Windows, version 17.6 (MedCalc Software, Ostend, Belgium) and Stata Statistical Software, release 14 (StataCorp. 2015. College Station, TX.).

3. RESULTS The mean age of the participants was 45.7 ± 13.7 years, with a mean BMI of 27.6 ± 4.7 kg/m2. Three subjects had an apnoea-hypopnoea index (AHI) of < 5, seven had mild OSA (5 ≤ AHI < 15), seven had moderate OSA (15 ≤ AHI < 30), four had severe OSA (AHI ≥ 30). Subject characteristics are summarised in Table 3.1. All subjects OSA subjects Male (n=15) Female (n=6) 14 Male 4 Female Age (years) 45.1 ± 14.7 47.0 ± 12.2 43.8 ± 13.5 BMI (kg/m2) 27.8 ± 4.4 27.2 ± 5.8 28.0 ± 4.8 Neck circumference (cm) 39.8 ± 2.9 36.1 ± 4.0 39.2 ± 3.4 AHI 23.4 ± 21.2 12.6 ± 14.3 23.4 ± 19.8 Angle  (º) 40.2 ± 1.9 41.2 ± 1.3 40.3 ± 1.8

Angle  (º) 78.5 ± 1.7 78.7 ± 2.0 78.7 ± 1.8 Pharyngeal length (mm) 56.4 ± 8.3 44.4 ± 7.1 53.2 ± 10.1 Tongue base angle (º) 64.9 ± 10.9 70.2 ± 11.5 66.3 ± 11.3 Upper pharyngeal CSA (mm2) 54.1 ± 19.6 43.1 ± 17.1 50.8 ± 19.9 Lower pharyngeal CSA (mm2) 245.4 ± 55.0 160.1 ± 70.7 225.6 ± 72.0 Soft palate length (mm) 37.0 ± 5.2 31.9 ± 3.9 35.7 ± 5.4 Table 3.1. Characteristics of 21 subjects (including 18 OSA subjects). Data are expressed as mean ± SD. Body mass index (BMI). Apnoea-hypopnoea index (AHI). Cross-sectional area (CSA). Angle  denotes angle between the horizontal plane and a line from the tragus to the spinous process of the C7 vertebrae. Angle  denotes angle between the horizontal plane and a line from lateral angle of eye to the tragus.

For ultrasound and MRI analysis, the location and characteristics of the grid are summarised in Table 3.2. On average, grid point 1 (lower left) of the anterior grid was ~21 mm from the internal mental spine of the mandible (insertion of genioglossus) under both MRI and ultrasound imaging, and ~30 mm for grid point 1 (lower left) of the posterior grid. Grid point 1 of the posterior column experiment was ~33 mm from the genioglossus insertion point. Ultrasound MRI

Anterior Distance of point 1 from internal 21.1 ± 3.6 21.1 ± 3.7 15-point mental spine of the mandible (mm) grid Distance between grid columns 2.9 ± 0.2 2.9 ± 0.3 (mm)

Distance between grid rows (mm) 2.0 ± 0.2 2.2 ± 0.3

Posterior Distance of point 1 from internal 28.3 ± 3.8 32.0 ± 4.7 15-point mental spine of the mandible (mm) grid Distance between grid columns 2.7 ± 0.4 2.9 ± 0.4 (mm)

Distance between grid rows (mm) 2.0 ± 0.1 2.3 ± 0.3

Posterior Distance of point 1 from internal 32.0 ± 4.4 33.9 ± 5.0

5-point mental spine of the mandible (mm) column Distance between grid columns 5.5 ± 0.7 5.5 ± 0.8

(mm)

Table 3.2. Image grid / column characteristics across experiments. Data are expressed as mean ± SD.

For ultrasound image analysis, the internal mental spine of the mandible denotes the point of insertion of genioglossus (refer to Supplementary Video to see MRI and ultrasound tracking of a typical subject’s genioglossus during a typical analysed breath).

Mean maximal Differences in measurement of mean maximal inspiratory inspiratory displacement of tongue (MRI – displacement of ultrasound) tongue Mean ± SD (95% limits of ICC (95% CI) (mm) agreement) (mm)

All subjects (n=21)

Anterior grid MRI: 0.37 ± 0.17 0.05 ± 0.21 (-0.35 to 0.46) 0.64 (0.54 to 0.72)

US: 0.32 ± 0.14

Posterior grid MRI: 0.97 ± 0.44 -0.24 ± 0.64 (-1.49 to 1.03) 0.63 (0.50 to 0.73)

US: 1.20 ± 0.66

Posterior MRI: 0.77 ± 0.41 -0.02 ± 0.58 (-1.15 to 1.11) 0.74 (0.61 to 0.82) column US: 0.79 ± 0.57

OSA subjects (n=18)

Anterior grid MRI: 0.35 ± 0.16 0.05 ± 0.11 (-0.35 to 0.46) 0.62 (0.51 to 0.71)

US: 0.30 ± 0.14

Posterior grid MRI: 0.93 ± 0.46 -0.24 ± 0.47 (-1.54 to 1.06) 0.66 (0.53 to 0.74)

US: 1.17 ± 0.71

Posterior MRI: 0.74 ± 0.40 -0.07 ± 0.58 (-1.16 to 1.02) 0.77 (0.65 to 0.85) column US: 0.81 ± 0.61

Table 3.3. Reproducibility of measurements. Reported data are derived from the mean maximal inspiratory tongue displacement of the regional grid/column points across 3 separate analysed breaths in 21 subjects unless specified. Results reported as mean ± SD. CI, confidence interval; ICC, intraclass correlation coefficient; MRI, magnetic resonance imaging; US, ultrasound.

Table 3.3 reports the reproducibility analysis, demonstrated by estimates of the mean difference, its 95% limit of agreement and the intraclass correlation coefficient. Overall,

there was good reproducibility of maximal inspiratory tongue movement in all measured regions between the 2 imaging modalities (Figs. 3.3 and 3.4), with larger displacement measured in the posterior tongue than anterior (Table 3.3). Good agreement and reproducibility were also reported in OSA subjects. Refer to Supplementary Videos 1-3 for inspiratory movement of posterior tongue grids measured with MRI and ultrasound for one representative subject.

3.1 Anterior grid For the anterior grid, on average the 15 grid points displaced by 0.37 ± 0.17 mm measured using MRI, and 0.32 ± 0.14 mm measured using ultrasound for the 21 subjects during inspiration. The ICC was 0.64 (95% CI: 0.54 to 0.72), with a mean difference of 0.05 ± 0.21 mm (95% limits of agreement: -0.35 to 0.46) (Figs. 3.3 and 3.4). Two-way ANOVA did not reveal any significant interaction between the maximal inspiratory displacement of the anterior, middle and posterior grid groups and the imaging modality (F2,120 = 0.249, p = 0.780).

3.2 Posterior grid For the posterior grid across 21 subjects, on average the 15 grid points displaced by 0.97 ± 0.44 mm measured using MRI, and 1.20 ± 0.66 mm measured using ultrasound during inspiration. The ICC was 0.63 (95% CI: 0.50 to 0.73), with a mean difference of -0.24 ± 0.64 mm (95% limits of agreement: -1.49 to 1.03) between MRI and ultrasound measurements, where ultrasound measured significantly more displacement than MRI (p = 0.003) (Figs. 3.3 and 3.4). Two-way ANOVA did not reveal any significant interaction between the maximal inspiratory displacement of the anterior, middle and posterior grid groups and the imaging modality (F2,120 = 0.016, p = 0.984) (Fig. 3.5). Combining the anterior and posterior tongue grids, the ICC between MRI and ultrasound was 0.79 (95% CI: 0.75 to 0.82) across the 21 subjects. To assess sample size calculation adequacy, a paired-samples t-test was conducted and we measured a significant difference for MRI (0.97 ± 0.44 mm) and ultrasound (1.20 ± 0.66 mm); with a mean difference of 0.24 ± 0.45 mm (95% CI: 0.03, 0.44); t(20)=2.42, p = 0.025.

Fig 3.3. Bland-Altman plots of the difference between ultrasound and magnetic resonance imaging in measuring tongue displacement. Plots of the difference between ultrasound and magnetic resonance imaging (MRI) in measuring the maximal inspiratory tongue displacement against their mean for (a) the posterior grid, (b) the anterior grid and (c) posterior column. Coloured markers indicated matching grid points of the same subject. The upper and lower lines indicate the 95% limits of agreement. Right column denotes results of 21 subjects, and the left column denotes results of 18 OSA subjects.

Fig 3.4. Scatter plots of the tongue grids and posterior column displacement between ultrasound and magnetic resonance imaging. Measurement in mm. MRI = Magnetic resonance imaging. Subject legend in middle of figure. Left column denotes results of all 21 subjects, right column denotes results of 18 OSA subjects. Blue line represents the regression line. The brown dotted line represents the 95% confidence interval for the regression line. The grey broken line represents the 95% prediction interval for the regression line. For any given value of the independent variable (MRI displacement), this interval represents the 95% probability for the values of the dependent variable (ultrasound displacement).

Fig 3.5. Scatter plots of the posterior tongue regional displacement between ultrasound and magnetic resonance imaging. Measurement in mm. MRI = Magnetic resonance imaging. Anterior = grid points 1-5, Middle = grid points 6-10, Posterior = grid points 11-15. Infero-posterior grid group (defined as points 11 and 12). Blue line represents the regression line. The brown dotted line represents the 95% confidence interval for the regression line. The orange broken line represents the 95% prediction interval for the regression line. For any given value of the independent variable (MRI displacement), this interval represents the 95% probability for the values of the dependent variable (ultrasound displacement).

3.3 Posterior column of 5 points In the single posterior column of 5 points, the points displaced on average by 0.77 ± 0.41 mm measured with MRI and 0.79 ± 0.57 mm measured with ultrasound during inspiration (see Fig. 3.6 for the peak inspiratory anterior and posterior tongue grids and posterior column displacements for the example subject depicted in Fig. 3.1). The ICC was 0.74 (95% CI: 0.61 to 0.82), with a mean difference of -0.02 ± 0.58 mm (95% limits of agreement: -1.15 to 1.11). A two-way ANOVA did not reveal any significant differences in maximal inspiratory

displacement of the 5 individual points between the imaging modalities (F4,200 = 1.886, p = 0.114). Simple main effect analysis did show that the more inferior points (1 and 2) displaced significantly more compared to the more superior points (4 and 5) (p < 0.001) during inspiration across both modalities (Table 3.4).

Fig 3.6. Peak inspiratory movement of anterior and posterior tongue grids and posterior column measured with MRI and ultrasound for one representative subject. Solid black circles denote the tracking grid points at the start of the respiratory cycle.

Ultrasonography measured more variability in the magnitude of posterior grid and column displacement across the 21 subjects compared to MRI (54.8% and 71.6% respectively for posterior grid and posterior column for ultrasound vs. 45.9% and 53.1% for MRI) (Table 3.5). The coefficient of variation was similar for both modalities in measuring maximal inspiratory displacement for the anterior tongue grid (45.1% vs. 44.7% respectively). However, there was a higher inter-breath variability (measured by mean inter-breath coefficient of variation) across the 21 subjects compared with the variability measured for each modality across all 3 measured regions (Table 3.5).

MRI Ultrasound 1 2 3 4 5 1 2 3 4 5 A 1.07 0.10 0.13 0.07 0.14 0.91 0.48 0.22 0.12 0.10 B 0.95 0.31 0.46 0.59 0.31 0.92 1.24 0.56 0.22 0.06 C 0.95 1.04 1.03 0.11 0.06 0.95 1.29 1.18 0.24 0.17 D 0.90 0.95 0.64 0.25 0.46 0.58 0.65 0.67 0.16 0.07 E 1.75 1.49 1.41 1.36 1.19 1.08 1.05 1.03 0.82 0.81 F 0.45 0.22 0.65 0.34 0.29 1.18 0.97 0.77 0.54 0.30 G 0.37 0.59 0.50 0.45 0.22 0.24 0.38 0.10 0.05 0.08 H 0.66 0.71 1.12 0.94 0.74 2.47 1.65 1.21 0.78 0.50 I 1.17 1.26 1.35 1.11 0.96 1.53 0.88 0.59 0.18 0.44 J 0.60 1.07 1.37 0.37 0.04 1.46 1.17 0.84 0.41 0.11 K 0.55 1.00 1.11 0.43 0.42 2.18 1.40 0.89 0.25 0.45 L 0.77 0.67 0.82 0.61 0.39 1.44 1.01 0.55 0.12 0.05 M 0.88 0.75 0.61 0.50 0.54 1.52 0.99 0.60 0.19 0.18 N 1.28 1.48 1.18 1.02 0.98 1.48 1.76 2.09 1.81 1.47 O 0.35 0.44 0.09 0.10 0.16 0.51 0.43 0.31 0.23 0.17 P 1.50 0.71 0.77 0.58 0.39 1.21 0.44 0.21 0.09 0.16 Q 0.78 2.12 2.48 0.85 0.88 2.22 2.43 2.01 0.90 1.66 R 0.60 0.15 0.41 0.36 0.23 0.10 0.14 0.09 0.07 0.11 S 1.95 1.89 1.35 0.74 0.74 3.92 2.24 1.32 1.43 2.09 T 1.92 1.24 1.46 1.49 1.49 1.51 1.09 0.36 0.27 0.06 U 0.26 0.55 0.28 0.35 0.22 0.28 0.13 0.05 0.09 0.01 Mean 0.94 0.89 0.92 0.60 0.52 1.32 1.04 0.74 0.42 0.43 ± SD ± ± ± ± ± ± ± ± ± ± 0.50 0.55 0.56 0.40 0.40 0.88 0.63 0.58 0.47 0.59 Table 3.4. Maximal inspiratory displacement of 5 posterior column points. Data are expressed as mean ± SD. Measurements are in mm. Subjects are denoted by alphabetical letters A to U. Posterior column points denoted from 1 to 5 (from inferior to superior).

Median Range Coefficient of variation (inter- (maximum, Across Mean of 3 Mean of 3 quartile minimum) modalities analysed analysed range) (mm) breaths breaths (mm) under under MRI ultrasound Anterior US: 0.26 US: 0.51 46.4% 71.7% grid (0.21 – 0.36) (0.67, 0.17) US: 44.7% (B1: 49.1%; (B1: 52.3%; MRI: 0.33 MRI: 0.59 MRI: 45.1% B2: 42.9%; B2: 70.8%; (0.29 – 0.50) (0.70, 0.17) B3: 47.2%) B3: 91.9%) Posterior US: 1.19 US: 2.71 55.9% 60.8% grid (0.74 -1.43) (2.81, 0.66) US: 54.8% (B1: 54.9%; (B1: 60.0%; MRI: 0.92 MRI: 1.41 MRI: 45.9% B2: 58.5%; B2: 60.2%; (0.66 – 1.49) (1.71, 0.30) B3: 54.3%) B3: 62.2%) Posterior US: 0.70 US: 2.10 73.1% 66.0% column (0.42 – 0.96) (2.20, 0.10) US: 71.6% (B1: 78.3%; (B1: 70.3%; MRI: 0.66 MRI: 1.29 MRI: 53.1% B2: 68.7%; B2: 63.3%; (0.42 – 1.17) (1.52, 0.23) B3: 72.2%) B3: 64.6%) Table 3.5. Variability of measured mean maximal inspiratory regional tongue displacement. Data are derived from the mean maximal inspiratory tongue displacement of the regional grid/column points across 3 separate analysed breaths in 21 subjects unless specified. Interquartile range reported as first quartile to third quartile. Mean coefficient of variation (CoV) of 3 analysed breaths is the mean CoV of B1, B2, B3 representing first, second and third analysed breaths under the respective image modality. MRI, magnetic resonance imaging; US, ultrasound.

4. DISCUSSION This is the first study to directly compare ultrasonography and MRI in measuring tongue motion during respiration. We demonstrated that our ultrasonography technique has good consistency and agreement with MRI in measuring maximal inspiratory displacement of the tongue in awake supine subjects with and without OSA. Ultrasound also recorded higher variance in the measured maximal inspiratory displacement in the posterior tongue, but little difference in this variability in the anterior tongue where less movement was recorded.

Differences between the two modalities are of similar or lower magnitude than inter-breath variation within a modality, indicating that ultrasound provides similar measure of tongue displacement to MRI.

In previous MRI and ultrasound studies, the maximal displacement in the tongue in healthy awake subjects during inspiration is observed within the infero-posterior region, of around 1mm (Cheng et al., 2008; Kwan et al., 2014; Kwan et al., 2015). This motion is more in the anterior direction than the caudal direction, and more than other parts of the tongue, consistent with the direction of the horizontal genioglossus fibers in this region. In this study involving thin and obese subjects with and without OSA, both imaging modalities confirmed this finding. The mean maximal inspiratory displacement of the posterior grid was ~ 1 mm, compared to ~ 0.4 mm for the anterior grid. Ultrasonography recorded more variability across the region compared to MRI, as well as larger inter-breath variability across the 3 measured breaths. This observed breath-to-breath variability within subjects is similar to previous MRI study (Brown et al., 2013a) and it possibly represents the delicate moment-to- moment coordination of pharyngeal dilator muscles activity and drive in order to maintain pharyngeal patency within and between breaths (Bilston and Gandevia, 2014). The larger variability recorded with ultrasound in this study could be due to the better temporal and spatial resolution of ultrasonography (40Hz frame rate for ultrasound vs 4Hz for MRI). Comparison of precise tidal volume measurements of each analysed breath were not possible in this study as MRI compatible inductive plethysmographs were not available. However, all analysed breaths were monitored through the digital output of the abdominal respiratory band and comparable consistent stable breaths were chosen. Our results recording more displacement in the posterior tongue during inspiration were also consistent with previous published MRI results (Brown et al., 2013a; Cheng et al., 2008) and ultrasound results (see Chapter 2).

4.1. MRI applications and limitations Currently MRI provides excellent upper airway and soft tissue resolution and accurate quantification of cross-sectional area and volume (e.g. Abbey et al., 1989; Ritter et al., 1999; Rodenstein et al., 1990; Ryan et al., 1991a; Shelton et al., 1993a; Shelton et al., 1993b), including 3D anatomical analysis and volume calculation of the tongue (e.g. Stone et al., 2016; Woo et al., 2015; Woo et al., 2012). Recent development in MR elastography further allows measurement of viscoelastic properties of soft tissues in-vivo (Brown et al., 2015;

Cheng et al., 2011b) and provides the ability to define a biomechanical model of the tongue and upper airway in healthy and patients with different respiratory disorders (e.g. Buchaillard et al., 2007; Cheng et al., 2011b; Gerard et al., 2005; Stavness et al., 2008). However, MRI has not been widely adopted as standard clinical tool for upper airway disorders due to difficulty in performing MRI in asleep subjects, limited availability, high cost, exclusion of patients suffering from claustrophobia or with MR incompatible implants. Under MRI, one technical difficulty is to ensure consistency in subjects’ breathing pattern during scanning. Long scan times, confined space, and the noisy environment in the scanner may cause increased restlessness which may increase breathing variability. Another limitation during MR tagging analysis is tracking the drift of grid point which can occur over a respiratory cycle, due to segment-to-segment tracking errors and/or breath-to-breath variability. The SPAMM technique is also limited in imaging an entire respiratory cycle, as tag fading occurs after ~1 s. This requires retagging and joining adjacent segments together, which have a ~140 ms gap between them due to the retagging.

4.2. Ultrasonography applications and limitations Compared to MRI, ultrasonography is non-invasive, provides real time imaging and has a lower cost. Previous studies demonstrated its ability to provide anatomical and dynamic measurement of upper airway muscular structures in OSA patients (Lahav et al., 2009; Shu et al., 2013). It has also been used to provide an approximate 3D measurement of the tongue volume from repeated 2D measurements across different planes of views (e.g. Flowers et al., 2005; Yang and Stone, 2002). Furthermore, ultrasound elastography, a technique first described in 1990s, has recently been used to provide qualitative and quantitative assessment of tissue stiffness (e.g. Shiina et al., 2015; for review, see Gennisson et al., 2013; Sigrist et al., 2017). It has also been used to measure mechanical properties of musculoskeletal tissue in clinical practice (Drakonaki, 2012; Drakonaki et al., 2012), and to diagnose tongue carcinoma in a preliminary study (Shingaki et al., 2016). A pilot study recently examined the use of ultrasound elastography to evaluate mechanical characteristics and textures of gels (Gao et al., 2016), which has a similar elastic property to that of the human tongue. Potentially ultrasonography could contribute to a dynamic clinical assessment of patients suspected or known to have OSA.

However, ultrasound imaging artifacts limit application of the technique (Kremkau and Taylor, 1986). Although air-tissue interfaces, such as the tongue surface, provide excellent

contrast due to high acoustic impedance differences, it is more difficult to visualise detailed intramuscular tongue architecture without significant increase in the contrast and/or resolution (Kremkau, 2016a). In the sagittal plane, precise definition of bony landmarks is also difficult due to high attenuation of ultrasound waves and may cause acoustic shadow that obscure part of the tongue image (Stone, 2005). Image quality can also vary between subjects due to tongue properties which may affect propagation of ultrasound wave. These include the amount of fat which causes scattering of soundwaves, and the interdigitate arrangement of intrinsic muscles of the tongue with the extrinsic muscles (Stal et al., 2003) which may increase refraction of soundwaves (Stone, 2005). Further developments in transducer and image analysis algorithms increase precision in analysing the motion, as well as grid position in relation to a bony reference point. In addition, ultrasound image acquisitions are sensitive to even small movements of both transducer and subject. This can be minimised by standardising the head position before and during ultrasonography, and ensuring the transducer is stabilised with intimate contact with the skin submentally at a constant pressure and position relative to the jaw.

Although image correlation analysis has good accuracy in other skeletal muscles (Chen et al., 1995), it has not been validated in human tongue muscles. Similar to MR tagging, apparent “drift” of the grid points can occur during analysis of ultrasound images. The analysis can also be affected by ultrasound storage system settings, where DICOM images are stored at default output video frame rate setting of 30 frames/s but images are collected at a scan frame rate of 70 Hz. This may cause underestimation of deformation (Koopman et al., 2010; Marwick, 2010) and may be inadequate for fast tongue motions and reduce its temporal resolution (Li et al., 2005). Despite these limitations, our technique has previously been shown to be reproducible within and across sessions on different days (see Chapter 2).

CONCLUSION This study demonstrated good agreement and consistency between MRI and ultrasound in measuring tongue movement in subjects with and without OSA. This indicates that ultrasound may be suitable for dynamic airway muscle function analyses, particularly when MRI is unavailable or too expensive. Future applications may provide further information and understanding of the biomechanics of upper airway muscles and their physiological responses in healthy and disease states.

CHAPTER 4

Influence of respiratory mechanics and drive on genioglossus movement under ultrasound imaging

SUMMARY

Genioglossus electromyographic activity has been reported to increase during inspiratory flow-resistive loading, during passive ventilation with increasing intrapharynegal negative pressure, and during voluntary hyperventilation in the supine position. The present study used ultrasound to assess genioglossus movement in conditions in which ventilatory drive or respiratory mechanics were changed. We measured genioglossus motion in 20 healthy awake subjects (21-38 years) in supine position (i) during spontaneous breathing, voluntary targeted breathing (normal tidal volume Vt), and voluntary hyperpnoea (at 1.5Vt and 2 Vt); (ii) during inspiratory flow resistive loading; (iii) with changes in end-expiratory lung volume (EELV). In healthy subjects, we observed non-uniform heterogeneous inspiratory motion (peak displacement between 0.5 – 2 mm) within the infero-posterior part of genioglossus during spontaneous quiet breathing, with more displacement in the posterior region than the anterior. A significantly increased mean peak inspiratory displacement was recorded with increased Vt (0.89±0.56 mm; 1.02±0.88 mm; 1.27±0.70 mm respectively for voluntary Vt, and during voluntary hyperpnoea at 1.5Vt and 2Vt), but the regional heterogeneity within the infero- posterior genioglossus disappeared during voluntary targeted breathing compared with quiet breathing. This may be due to different neural drive to genioglossus during voluntary breathing. During increasing inspiratory resistive loading, the genioglossus displaced less anteriorly and more inferiorly. The observed motion may serve to maintain upper airway patency by balancing intraluminal negative pressure with positive pressure generated by upper airway dilatory muscles. When lung volume was altered, no significant changes in genioglossus movement were observed. Our findings provide improved understanding and quantification of genioglossus movement during conditions where ventilatory drive is altered.

1. INTRODUCTION The human upper airway is important in many voluntary and involuntary tasks such as swallowing, speech and breathing. Patency of the upper airway requires a dynamic coordinated system that can rapidly change dilator muscle activity to counterbalance pressures that act to collapse it (Bilston and Gandevia, 2014; Davidson, 2003). The system varies with the respiratory cycle, with changes in head positions and posture, and with the state of wakefulness and sleep (e.g. Douglas et al., 1993; Liistro et al., 1988; Mezzanotte et al., 1992; Remmers et al., 1978; Spann and Hyatt, 1971; Suratt et al., 1988; Yildirim et al., 1991). Passive mechanical properties of the upper airway and surrounding tissues determine its propensity to move and collapse (Bilston and Gandevia, 2014; Fregosi and Ludlow, 2014). The system is also influenced by active moment-to-moment changes such as pressures generated as a result of neural drive to dilator and other upper airway muscles (Fregosi and Ludlow, 2014). The largest upper airway dilator is genioglossus (Mu and Sanders, 2000). Neural drive to the human genioglossus increases in hypercapnic and hypoxic conditions (Jordan et al., 2002; Klawe and Tafil-Klawe, 2003; Mateika et al., 1999; Onal et al., 1981; Parisi et al., 1987; Patrick et al., 1982), with a parallel increase to the hyoglossus (Mateika et al., 1999). Genioglossus electromyographic (EMG) activity has been reported to increase during inspiratory flow-resistive loading (Malhotra et al., 2000a; Pillar et al., 2001). Fogel and colleagues similarly found that phasic genioglossus EMG increases with increasing intrapharynegal negative pressure during passive ventilation, and inspiratory resistive loading (Fogel et al., 2001). During voluntary hyperventilation in the supine position, genioglossus EMG increases when compared to tidal breathing at rest (Eastwood et al., 2003; Vranish and Bailey, 2015).

Genioglossus movement during quiet breathing has been measured with magnetic resonance imaging (MRI) (Brown et al., 2013a; Cheng et al., 2008, 2011a) and ultrasound imaging (Kwan et al., 2014; Kwan et al., 2015). Other imaging techniques have been used to examine the morphology and mechanical behaviour of the soft tissues in the upper airway, but are limited for quantifying genioglossus movement. MRI provides excellent upper airway and soft tissue resolution (Ryan et al., 1991a; Schwab, 1996, 1998). Tagged MRI studies revealed maximal anterior movement in the infero-posterior region of genioglossus of about 0.5 - 2 mm during tidal inspiration and posterior movement during expiration (Brown et al., 2013a; Cheng et al., 2008, 2011a). In healthy awake subjects, MRI recorded reduced

overall anterior movement of genioglossus, axial plane cross-sectional airway area, and mid- sagittal plane anteroposterior airway diameter during inspiratory resistive loading (Cheng et al., 2011a; How et al., 2007). This is likely due to a shift in ability of upper airway dilator muscles to counteract the increased negative intraluminal pressure. Limitations of MRI include low temporal resolution, high cost, noise and limited availability.

Ultrasonography has been increasingly used to image the upper airway in real time in both experimental and clinical settings (e.g. Hamlet and Reid, 1972; Kelsey et al., 1969; Kundra et al., 2011; Prasad et al., 2011; Stone, 2005; Sustic, 2007; Tsui and Hui, 2008; Wojtczak, 2012). In Chapter 2 we reported a novel ultrasound research technique to measure genioglossus movement during quiet breathing in healthy subjects. It has good consistency and reliability and showed genioglossus motion beginning before inspiratory airflow, with ~ 1 mm (predominantly anterior) displacement during inspiration. The technique revealed further heterogeneous (non-uniform) inspiratory motion within the postero-inferior genioglossus, with the posterior region displaced more than the anterior region.

Previous EMG studies confirmed increased genioglossus EMG in conditions with increased inspiratory neural drive, but quantitative measurement of regional genioglossus movement during inspiration in awake subjects is lacking. Therefore, to understand how altered respiratory mechanics and neural drive influence the physiological behaviour of genioglossus, the present study was designed to measure its movement using ultrasonography. We hypothesised that increased genioglossus inspiratory movement would occur in three selected physiological conditions in which inspiratory neural drive is increased: (i) during voluntary hyperpnoea, (ii) during inspiratory flow resistive loading, and (iii) with large changes in the end-expiratory lung volume (EELV), although anterior genioglossus movement would be counteracted by the forces produced by any changes in upper airway negative pressure.

2. METHODS Subjects with history of an active or chronic respiratory or sleep disorder and those using medication that could affect respiration or ventilation were excluded from the study. Thirty- eight healthy subjects (19 males and 19 females) in total were recruited across the three experiments. For each experiment, we recorded data from twenty (10 males and 10 females)

subjects. Some subjects participated in two or all three experiments. All participants completed the Epworth Sleepiness Scale (Johns, 1991, 2002) and Berlin Sleep Questionnaire (Netzer et al., 1999), two self-report questionnaires measuring subjective daytime sleepiness as predictive tools for obstructive sleep apnoea (Abrishami et al., 2010; Pataka et al., 2014). Ethics approval was granted by the Human Research Ethics Committees of the Northern Sector of South East Sydney and Illawarra Area Health Service and University of New South Wales. The study was conducted according to the Declaration of Helsinki (2008) and informed written consent was obtained.

2.1 Experimental protocol The experimental setup is shown in Figure 4.1, and is similar to our previously reported method in Chapter 2. Head position was standardised with the Frankfort plane (defined by the inferior borders of the bony orbits and the upper margin of the auditory meatus) perpendicular to the horizontal bed surface to minimise variation in upper airway size due to flexion or extension of the head (Amis et al., 1999; Cai et al., 2016). The antero-posterior position of the head relative to the lower cervical spine was also standardised. The angle between horizontal plane and a line from tragus to spinous process of C7 vertebrae was constrained to 37 – 42°, and the angle between horizontal plane and a line from lateral angle of eye to tragus to 77 - 82°. Angle measurements were made with two goniometers, aligned along the measurement plane and the horizontal plane. Padding was used for the head, neck and shoulders to maintain the standardised posture. Head position was checked before capturing each image sequence to ensure no change in the measured angles from the start of the experiment. Subjects were asked to relax, remain awake, keep their mouth closed and to place the tongue in its usual ‘resting’ position, usually with its tip on the incisors. This tongue position has been described as close to the optimum genioglossus length for protrusion force (Saboisky et al., 2013; Sha et al., 2000). Subjects breathed through their nose throughout the study. The resting ventilation of each subject was monitored using calibrated respiratory inductance bands (Inductotrace, Ambulatory Monitoring Inc., Ardsley, New York, USA) over thorax and abdomen. Calibration was achieved with 800 mL bags. Real-time signals were digitised at 1 kHz using a CED1401 data acquisition system and Spike2 software (Cambridge Electronic Design, Cambridge, UK). Respiratory data were analysed off-line to determine the inspiratory time, tidal volume and respiratory rate. The onset of inspiration was taken from the signal of the abdominal inductance band.

Figure 4.1. Setup for each of the 3 conditions. Ultrasound transducer is positioned submentally. (a) Voluntary hyperpnoea experiment. Subject lay supine and directly viewed the oscilloscope screen which displayed a signal of volume and a “target tracking” waveform based on the subject’s average tidal volume and respiratory frequency. The subject followed this waveform in real time. (b) Inspiratory resistive loading experiment. The subject wore a modified nose mask and were requested to breathe exclusively through the nose. The breathing apparatus was connected to a low resistive 2-way valve so that inspiration occurred through a pneumotachometer and pressure transducer whilst expiration took place at the valve to minimise rebreathing. The pressure, flow and respiratory inductance band signals were recorded in real time. An inspiratory resistive load was added by restriction of airflow distal to the pneumotachometer. (c) Imposed changes in end-expiratory level. The subject was supine in a head-out rigid-shell ventilation chamber enclosing the whole body (truncated Figure). The chamber rested across the upper anterior chest just caudal to the suprasternal notch, with memory foam used to prevent pressure leak between chamber and torso. A vacuum/blower attachment was attached to the caudal part of the shell to allow changes in extra-thoracic pressure within the chamber.

2.2 Ultrasound scanning and analysis In Chapter 2, our ultrasound protocol has been shown to reliably image movement of the genioglossus during quiet breathing. In brief, ultrasound images were collected using a Philips iU22 system (Andover, USA) with a curved array C8-5 transducer, which has a probe frequency of 5 to 8 MHz. The handheld transducer was positioned submentally, and aligned in the mid-sagittal plane and pointed cranially. This provided a lateral view of the tongue body, submental musculature and mandible. Time gain compensation, depth and near gain control were adjusted manually to obtain the best image quality (refer to Fig 4.2 for

representative ultrasound image). The depth of the image acquisition was set to 6 cm. Output from the abdominal inductance band was recorded concurrently by the ultrasound machine, appearing as a waveform onscreen. This allows the synchronisation of genioglossus motion with respiration. A scan of the surrounding tissues was also performed prior to recording to determine the position of the superior surface of the tongue, the mandible and the posterior portion of the tongue. Real-time B-mode images were collected for at least 5 consecutive stable breaths at a frame rate of ~40 Hz (termed a “sequence”). A minimum of three sequences were captured in each imaging session.

Fig 4.2. Schematic representation of ultrasound image, grid position and numbering system. (a) Ultrasound image of the placement of grid with red line outlining the genioglossus and 15 solid black circles denoting the tracking points. GG – genioglossus, M – mucosa, S – tongue surface, GH – geniohyoid, MH – myohyoid (b) 15 grid points and groups represented. Points 1 – 5 are defined as the “anterior” group, 6 – 10 as the “middle” group and 11 – 15 as the “posterior” group. Points 11, 12 are defined as the “infero-posterior” group.

Image sequences were analysed off-line using custom image correlation software developed in MATLAB (Mathworks, MA, USA). For regions of suitable image quality, a rectangular grid measuring around 100mm2 area containing 3 columns (≈ 4 mm apart) of 5 points (≈ 3 mm apart) was then placed over the tongue. Grid point 1 was positioned approximately 2 cm posterior to and 1 cm rostral to the internal spine of the mandible, where maximal inspiratory displacement was demonstrated in our previous study in Chapter 2. The nomenclature for the grid points is given in Figure 4.2. For analysis, points 1 – 5 are defined

as the “anterior” group, 6 – 10 as the “middle” group and 11 – 15 as the “posterior” group. Points 11 and 12 are defined as the “infero-posterior” group. We excluded any sequence in which swallowing or jaw motion occurred. We then selected the sequence with the clearest images. A custom image correlation program tracked the movement of these markers throughout the video sequence (Bonnefous and Pesque, 1986; Dilley et al., 2001). In our protocol, regional displacements over the three most stable breaths in each sequence were analysed and the average resultant was calculated. Reported results in each experiment are the mean maximal inspiratory regional genioglossus displacement across 3 breaths for 20 subjects.

2.3 Genioglossus motion during voluntary hyperpnoea For each subject, the calibrated volume signal derived from the summed output of the two respiratory inductance bands was displayed on a digital oscilloscope (Tektronic TDS3014, Beaverton, Oregon, USA) directly above the subject within their field of view (Fig 4.1). Subjects first performed spontaneous quiet breathing without visual feedback. Three sequences during tidal breathing were then captured, termed “spontaneous tidal breathing”. The tidal volumes during these sequences were averaged. We then generated a “target” waveform based on average tidal volume and respiratory frequency. This was the “baseline” target-breathing pattern. A “tracking” waveform was then generated and the subject then practiced matching their real-time volume signal to the target waveform on screen. Once the lines were matched reasonably, 3 breathing sequences were captured, termed “voluntary tidal breathing” (Vt). Next, the participant tracked a signal that was 150% (1.5Vt) and 200% (2Vt) of the initial averaged tidal volume in random order, while maintaining the same breathing frequency. Subjects rested for 10 mins between each condition. During all tests, subjects were monitored to ensure they remained calm and alert.

2.4 Genioglossus motion with inspiratory resistive loading Subjects breathed through a tight-fitting nose mask. The mask was connected to a low- resistance, two-way valve, so that inspiration occurred through a pneumotachometer (model 3813, Hans Rudolph Inc, Kansas City, USA) and pressure transducer (DP45-16, Validyne Engineering Corp., Northbridge, USA) (Fig.4.1). The inspiratory line was connected to the pneumotachometer and pressure transducer via 50-mm and 3-mm diameter stiff tubes, respectively. Inspiratory resistive loads were applied by placement of 3 different calibrated cylindrical plastic blocks (30 mm long), each with a central hole of different diameter into the

inspiratory line to restrict airflow distal to the pneumotachometer. These added an inspiratory resistance of 11.6 cmH2O/l/s, 22.3 cmH2O/l/s, and 75.2 cmH2O/l/s, loads A, B and C, respectively. Ultrasound recording of genioglossus movement was performed once steady-state breathing was achieved.

2.5 Genioglossus motion with imposed changes in end-expiratory lung volume Subjects rested supine in a custom-made head-out rigid-shell external ventilator chamber. The chamber was made of polymethyl methacrylate on a light reinforced stainless steel skeleton. Memory foam was used to prevent a pressure leak at the head end, and placed over the second and third ribs anteriorly. A vacuum/blower attachment was connected to the caudal part of the shell, to change extra-thoracic pressure and thus lung volume (Fig 4.1). Extra-thoracic pressure was measured inside the chamber via a pressure transducer (DP45-16, Validyne Engineering Corp., Northbridge, USA). Changes in end expiratory lung volume (EELV) were measured with the calibrated inductance bands (see above Section 2.1). Chamber pressure was adjusted to increase EELV by 1 L or to decrease it by 0.5 L from baseline, in random order. Lung volume changes were assessed by changes in the EELV indicated by the respiratory inductance bands. Ultrasound recording of genioglossus movement was performed once a stable end expiratory level was achieved.

Once analysis of this study was completed, the maximal resultant inspiratory displacement of the 15 grid points was less than in other experiments reported here (see Results) as well as our previous reports in Chapters 2 and 3. Therefore, a further study was performed with 6 subjects (3 males and 3 females) randomly selected from the original set of 20 subjects to determine if the presence of the ventilatory chamber across the anterior chest wall contributed to the reduced movement. Six sequences of tidal breathing were collected, half with the ventilator shell positioned over the chest and body in a random order. Ultrasound recording of genioglossus movement was performed in the usual way once stable breathing was achieved.

2.6 Statistical analysis Means and standard deviation (SD) were used for descriptive purposes. To assess gender differences in the subject characteristics we used independent sample t-tests. To assess differences in maximal displacement in different genioglossus regions across different

conditions, a one-way ANOVA with repeated measures with a Greenhouse-Geisser correction was used. Two-way ANOVA analysis was used to compare the effect of gender on the magnitude of regional GG displacement. Post-hoc analyses were performed with Bonferroni’s correction, or Tukey’s test for pairwise comparison. Statistical analyses were carried out using SPSS version 23.0 (Armonk, NY, USA). Statistical significance was accepted at p<0.05.

3. RESULTS Subject characteristics are shown in Table 4.1. In each experiment, there was no difference between the genders for age, Epworth sleepiness score and Berlin questionnaire, but females had a smaller neck circumference than males (31.8  1.7 cm vs. 37.6  2.4 cm; mean  SD) (t(38) = 8.622, p<0.001), and lower BMI (21.0  2.0 vs. 23.2  2.9 kg/m2) (t(38) = 2.695, p=0.011).

Male (n=19) Female (n=19) p value

Age (years) 30.4 ± 4.9, (21 – 38) 27.6 ± 4.1, (21 – 37) 0.060

BMI (kg/m2) 23.2 ± 2.9, (17.3 – 27.4) 21.0 ± 2.0, (17.5 – 24.5) 0.011

Neck circumference (cm) 37.6 ± 2.4, (35 – 41) 31.8 ± 1.7, (28 – 34) <0.001

Epworth sleepiness score 3.1 ± 1.7, (1 – 7) 2.4 ± 1.6, (1 – 7) 0.250 (range 1-24)

Berlin questionnaire Low Low

Table 4.1. Characteristics of the total pool of 38 subjects. Data are expressed as mean ± SD with the range given in brackets. Body mass index (BMI). An Epworth sleepiness score of 11 - 24 is indicative of increased daytime sleepiness (Johns, 2002). The Berlin questionnaire has 3 categories related to the risk of sleep apnoea: “high risk” if there are 2 or more categories where the score is positive, and “low risk” if there is only 1 or no categories where the score is positive (Netzer et al., 1999).

The location and characteristics of the grid positioned over genioglossus, and subjects’ head positions are shown in Table 4.2 for each experimental condition (see Fig 4.2 and Methods for definition of grid points). Data in Supplementary Tables 1, 2 and 3 reports the

mean maximal inspiratory displacement of each of the 15 grid points during all 3 experiment conditions for 20 subjects. Supplementary Tables are included after Conclusion. Voluntary Lung volume Inspiratory hyperpnoea alteration resistive load

Mean posterior distance of 23.5 ± 4.4 17.7 ± 5.0 22.4 ± 5.0 point 1 from internal mental spine of the mandible (mm)

Mean rostral distance of 13.2 ± 2.7 12.7 ± 3.5 11.8 ± 2.4 point 1 from internal mental spine of the mandible (mm)

Mean distance between 3.3 ± 0.3 4.4 ± 0.5 3.3 ± 0.4 grid columns (mm)

Mean distance between 2.5 ± 0.2 3.1 ± 0.4 2.3 ± 0.3 grid rows (mm)

Mean angle  () 78.4 ± 2.9 78.4 ± 1.9 80.8 ± 1.8

Mean angle  () 38.8 ± 2.7 39.4 ± 2.1 40.5 ± 2.1

Table 4.2. Image grid characteristics across experiments. Data are expressed as mean ± SD. For image analysis, the internal mental spine of the mandible denotes the point of insertion of genioglossus). Angle  denotes angle between the horizontal plane and a line from the tragus to the spinous process of the C7 vertebrae. Angle  denotes angle between the horizontal plane and a line from lateral angle of eye to the tragus.

3.1 Voluntary hyperpnoea For the 20 subjects, the mean respiratory rate was not significantly different between the 4 experimental conditions; spontaneous breathing, targeted voluntary to Vt, 1.5Vt, and 2Vt. The mean tidal volume increased close to the expected target of 50% and 100% above baseline spontaneous breathing tidal volume (Table 4.3). During spontaneous breathing, there was overt anterior movement of the genioglossus as reported previously in Chapters 2 and 3. On average (across all 15 points of the grid) the anterior movement was 0.88 ± 0.13 mm, and the overall resultant measured 1.03 ± 0.51 mm (Table 4.4, S1 Table).

Spontaneous “Target” 1.5x 2x p tidal baseline “Target” “Target” value# breathing tidal baseline baseline breathing

Tidal volume (mL) 394 ± 161 419 ± 165 625 ± 265 752 ± 264 <0.001

Respiratory rate 13.9 ± 4.7 13.9 ± 4.6 14.2 ± 5.0 14.1 ± 4.9 0.510 (breaths / min)

Respiratory cycle 4.88 ± 1.96 4.86 ± 2.02 4.82 ± 2.10 4.84 ± 0.072 length (sec) 2.07

Inspiratory time 1.79 ± 0.90 2.01 ± 0.88 2.02 ± 0.90 2.06 ± <0.001 (sec) 1.03

Table 4.3. Respiratory variables for the voluntary hyperpnoea experiment. Data are expressed as mean ± SD. There was no significant difference between the mean tidal volumes during spontaneous tidal breathing and “target” baseline tidal breathing experiments across 20 subjects (p=0.06). There was no significant difference between the mean inspiratory time during all 3 “Target” breathing experiments across 20 subjects (p = 1.000) # p value denotes significance of results comparing between the 3 “Target” breathing experiments.

During voluntary breathing, the mean resultant peak displacement of the infero-posterior grid increased with increasing tidal volume to 0.89 ± 0.56 mm, 1.02 ± 0.88 mm and 1.27 ± 0.70 mm respectively for the “target tidal”, “1.5Vt” and “2Vt” conditions (Table 4.4, see Fig 4.3 for mean inspiratory movement of the grid). A significant difference was observed in the magnitude of peak resultant movement between the 3 “target breathing” conditions (ANOVA with repeated measures with a Greenhouse-Geisser correction, F1.746,33.168 = 4.488, p=0.023). There were non-significant changes in resultant displacement when subjects increased their targeted tidal volume by 1.5 times from “target tidal” baseline (1.02 ± 0.88 mm vs. 0.89 ± 0.56 mm, respectively, p = 0.483) or when targeted tidal volume increased from 1.5 times baseline Vt to 2 times baseline Vt (1.02 ± 0.88 mm vs. 1.27 ± 0.70 mm, respectively, p = 0.577). However, there was a statistically significant difference in the magnitude of the peak resultant displacement between “target tidal” and 2 times baseline Vt (p = 0.007). There was no significant interaction between gender and voluntary Vt on the maximal resultant displacement across the 15 grid point (two-way ANOVA F2,54 = 0.033, p=0.967).

Mean Anterior Middle Posterior Infero- across 15 group group group posterior points group

Spontaneous 1.03 ± 0.51 0.93 ± 0.49 1.00 ± 0.52 1.16 ± 0.56 1.13 ± 0.54 tidal

“Target” 0.83 ± 0.56 0.84 ± 0.56 0.83 ± 0.60 0.85 ± 0.55 0.89 ± 0.56 baseline tidal

1.5 x “Target” 1.07 ± 0.89 1.08 ± 0.92 1.14 ± 0.95 1.00 ± 0.81 1.02 ± 0.88

2x “Target” 1.24 ± 0.59 1.24 ± 0.58 1.29 ± 0.60 1.21 ± 0.61 1.27 ± 0.70

p value# 0.042 0.055 0.028 0.034 0.081

Table 4.4. Average displacement of infero-posterior region of genioglossus (voluntary hyperpnoea). Mean maximal inspiratory displacement (mm) of different groups within infero-posterior genioglossus grid for 20 subjects. See Methods for grid group definition. Data are expressed as mean ± SD. # p value denotes significance of results comparing between the 3 “Target” breathing experiments.

Further analysis demonstrated significant regional variation in the magnitude of the peak resultant movement within the focused infero-posterior part of genioglossus. During “spontaneous tidal” breathing, the posterior grid group moved 16% and 25% more than the middle or anterior groups respectively (F1.275,24.22 = 16.484, p<0.001) (Table 4.5). Similar to our previous finding in Chapter 2, maximal displacement was recorded for the most infero- posterior grid group (11,12) measuring 1.13 ± 0.54 mm. In contrast, during the 3 voluntary breathing conditions, this regional variation was lost, with all 3 grid groups moving in a more uniform “en-bloc” pattern (F1.451,27.578 = 0.073, p=0.874).

100

Fig 4.3. Mean peak inspiratory movement of grid during voluntary hyperpnoea experiment. Mean for the 15 grid points at the start of the respiratory cycle are denoted as solid black circles. Fig 4.3(a), (c) Spontaneous and targeted tidal breathing. Fig 4.3(b), (d) During voluntary study with two volumes. Fig 4.3(a), (b) Mean for 20 subjects. Fig 4.3(c), (d) Result for one of the study subject.

3.2 Inspiratory resistive loading When the 20 subjects inspired against an external inspiratory resistive load, inspiratory time and tidal volume increased with increasing resistance (Table 4.5). Compared to spontaneous tidal breathing, tidal volume increased by around 23% when subjects inspired against the highest resistance of 75.2 cmH2O/L/s. During spontaneous breathing, anterior movement of the genioglossus again occurred. The mean anterior displacement of the grid across 20 subjects was 0.73 ± 0.12 mm and the mean resultant movement was 0.89 ± 0.43 mm. Table 4.6 shows the mean resultant across the 15 points, and for the anterior, middle

101

and posterior grid groups within the infero-posterior part of genioglossus. Mean peak inspiratory movement of the grid during different inspiratory resistive loading is shown in Fig 4.4. Spontaneous Load A Load B Load C p value# tidal breathing

Tidal volume (mL) 493 ± 169 510 ± 140 510 ± 157 605 ± 266

Respiratory rate 13.3 ± 3.5 13.0 ± 3.9 12.8 ± 3.7 9.7 ± 4.0 <0.001

(breaths / min)

Respiratory cycle 4.94 ± 1.84 5.20 ± 5.09 ± 7.23 ± <0.001 length (sec) 2.36 1.51 3.00

Inspiratory time (sec) 1.30 ± 0.49 1.44 ± 1.49 ± 2.02 ± 0.093 0.35 0.57 1.42

Borg score 1.0 ± 1.0 2.5 ± 1.45 3.7 ± 1.73 6.0 ± 1.63 <0.001

Table 4.5. Respiratory variables for the inspiratory resistive load experiment. Data are expressed as mean ± SD. Load A, B and C added an inspiratory resistance of 11.6 cmH2O/L/s, 22.3 cmH2O/L/s, and 75.2 cmH2O/L/s respectively. The mean respiratory rate was significantly lower only during inspiratory resistive loading (experiment C). # p value denotes significance of results comparing between the 4 “breathing experiments”.

With increasing inspiratory resistive load, maximal anterior displacement of the grid points decreased progressively. This was statistically significant (F1.623,30.841 = 7.017, p=0.005). A Tukey post-hoc test revealed this maximal anterior displacement was significantly less with load “A” (0.47 ± 0.40 mm, p=0.009), load “B” (0.30 ± 0.59 mm, p<0.001) and load “C” (0.08 ± 0.81 mm, p=0.02) compared to spontaneous breathing without a resistive load (0.73 ± 0.12 mm) (Table 4.6, S2 Table). Maximal inferior displacement increased with increased inspiratory resistive load. It averaged 0.53 ± 0.33 mm, 0.57 ± 0.47, and 0.85 ± 0.62 mm along the coronal plane with load “A”, “B” and “C” respectively (Table

4.6, S2 Table). The result was also statistically significant (F2.211,42.005 = 3.777, p=0.027). No gender influences on the magnitude of maximal inspiratory resultant displacements were observed (F3,72 = 1.605, p=0.196).

102

Mean Anterior Middle Posterior Infero- across 15 group group group posterior points group

Spontaneous 0.89 ± 0.43 0.82 ± 0.38 0.94 ± 0.48 0.92 ± 0.47 0.97 ± 0.54 tidal Load A 0.78 ± 0.38 0.77 ± 0.45 0.78 ± 0.38 0.81 ± 0.40 0.79 ± 0.36

Load B 0.82 ± 0.55 0.77 ± 0.50 0.84 ± 0.59 0.89 ± 0.57 0.95 ± 0.55

Load C 1.18 ± 0.57 1.07 ± 0.52 1.24 ± 0.59 1.26 ± 0.64 1.32 ± 0.71

p value# 0.043 0.102 0.030 0.031 0.024

Table 4.6. Average displacement of infero-posterior region of genioglossus (inspiratory resistive load). Mean maximal inspiratory displacement (mm) of different grid groups within genioglossus for 20 subjects. See Methods for grid group definition. Data are expressed as mean ± SD. Load A, B and C added an inspiratory resistance of 11.6 cmH2O/L/s, 22.3 cmH2O/L/s, and 75.2 cmH2O/L/s respectively. # p value denotes significance of results comparing between the 4 “breathing experiments”.

Fig 4.4. Mean peak inspiratory movement of grid for 20 subjects during resistive inspiratory load experiment and lung volume alteration experiment. Mean for the 15 grid points at the start of the respiratory cycle are denoted as solid black circles. (a) During inspiratory resistive loading experiment. Load A, B and C added an inspiratory resistance of 11.6 cmH2O/L/s, 22.3 cmH2O/L/s, and 75.2 cmH2O/L/s respectively. (b) During passive change in EELV experiment.

103

We again recorded significant non-uniform inspiratory motion within the infero-posterior genioglossus, with the posterior group moving more than the anterior or middle groups during spontaneous breathing (F1.735,30.558 = 4.165, p=0.029), with inspiratory resistive load

“B” (F1.99,34.401 = 4.651, p=0.016), and with inspiratory resistive load “C” (F1.976,37.538 = 10.748, p<0.0001), but not during inspiratory resistive load “A” (p=0.64) (Table 4.6). Maximal displacement was recorded for the most infero-posterior grid group, measuring 0.97 ± 0.54 mm during tidal breathing.

3.3 Change in lung volume produced by an external ventilator

During spontaneous breathing at a normal EELV with subjects lying supine in an external ventilator chamber, overt anterior genioglossus movement was again observed. The mean anterior movement for the 20 participants across all grid points was 0.35 ± 0.11 mm, with a mean resultant displacement of 0.39 ± 0.21 mm (Table 4.7 and S3 Table for 20 subjects and also for comparable displacements for 6 subjects inside and outside the ventilator chamber). Mean peak inspiratory movement of the grid during lung volume alteration is shown in Fig 4.4.

When positive extrathoracic pressure was applied (15.7 ± 4.2 cmH2O), the end expiratory lung volume decreased by 532 ± 159 mL, with displacement of the measured grid points of 0.48 ± 0.29 mm during inspiration. When negative extrathoracic pressure was applied (16.6 ±

4.1 cmH2O), the end expiratory lung volume increased by 1067 ± 191 mL, and the average maximal displacement across all 15 grid points measured 0.33 ± 0.25 mm (Table 4.7, Fig 4.4).

The displacements were not statistically significantly different (F1.486,28.243 = 2.474, p=0.115).

There was no significant gender influence on the magnitude of displacement (F2,54 = 0.185, p=0.831).

Significant regional variation in the magnitude of the peak resultant motion during inspiration was again recorded, with the posterior group of the grid displaced more during spontaneous breathing (F1.128,21.433 = 13.503, p=0.001), decreased EELV (F1.509,28.667 = 6.125, p=0.01), and increased EELV (F1.939,36.842 = 8.048, p=0.001) (Table 4.7). Maximal displacement was again recorded for the most infero-posterior group, measuring 0.62 ± 0.37 mm. In the further study with 6 subjects (3 males and 3 females) to determine if the presence of the ventilatory chamber across the anterior chest wall contributed to the reduced

104

movement, the respiratory parameters were similar when subjects were inside or outside the ventilator chamber (S4 Table).

Mean Anterior Middle Posterior Infero- across 15 group group group posterior points group

Spontaneous tidal 0.39 ± 0.21 0.32 ± 0.16 0.41 ± 0.22 0.52 ± 0.28 0.62 ± 0.37

Negative extra- 0.33 ± 0.25 0.27 ± 0.31 0.31 ± 0.20 0.43 ± 0.30 0.49 ± 0.33 thoracic pressure

Positive extra- 0.48 ± 0.29 0.40 ± 0.23 0.48 ± 0.31 0.57 ± 0.38 0.73 ± 0.46 thoracic pressure p value 0.115 0.169 0.066 0.263 0.064

6 subjects outside 0.81 ± 0.68 0.83 ± 0.71 0.86 ± 0.66 0.86 ± 0.62 0.97 ± 0.69 chamber

6 subjects inside 0.49 ± 0.58 0.56 ± 0.61 0.56 ± 0.54 0.58 ± 0.52 0.62 ± 0.53 chamber p value 0.011 0.038 0.037 0.029 0.033

Table 4.7. Average displacement of genioglossus (lung volume alteration). Mean maximal inspiratory displacement (mm) of different grid columns within genioglossus for 20 subjects. See Methods for grid group definition. Data are expressed as mean ± SD.

Fig 4.5. Mean peak inspiratory movement of grid for 6 subjects inside or outside chamber during lung volume alteration experiment. Mean for the 15 grid points at the start of the respiratory cycle are denoted as solid black circles.

105

Maximal inspiratory displacement of all grid groups within the infero-posterior genioglossus significantly decreased when subjects were inside the ventilator chamber (Fig 4.5, lower part of Table 4.7, S3 Table). The mean resultant peak displacement of the grid points was 0.81 ± 0.68 mm when the 6 subjects were outside the chamber and 0.49 ± 0.58 mm when subjects were inside the chamber (F1,5 = 15.579, p=0.011).

4. DISCUSSION Using our ultrasound technique described in Chapter 2, this study reports how much genioglossus displacement changed when ventilatory drive was altered. When subjects voluntarily adopted a higher tidal volume during the voluntary hyperpnoea conditions, genioglossus displacement increased. With inspiration against an increasing resistive load, tidal volume and inspiratory time increased, and the genioglossus displaced less anteriorly, but more inferiorly. When lung volume was altered with the external ventilator chamber, no significant changes in genioglossus movement were observed. In the substudy however, genioglossus displacement was significantly decreased when the six subjects were inside the chamber, suggesting an influence from the occlusive foam placed across the upper anterior chest wall. Similar to previous MRI and ultrasound studies, we recorded non-uniform inspiratory motion within the infero-posterior part of genioglossus during quiet breathing with mean peak displacement between 0.5 – 2 mm, and more displacement in the posterior region than the anterior (Cheng et al., 2008; Kwan et al., 2014; Kwan et al., 2015). However, this usual regional difference in motion was lost during voluntary targeted breathing, instead a more uniform antero-inferior “en-bloc” motion was observed. This suggests a potential role for cortical inputs during voluntary breathing, resulting in increased displacement in the anterior part of genioglossus. Table 4.8 summarises our ultrasound findings and previous reported MRI and genioglossus EMG findings.

With ultrasonography, we recorded a decrease in regional genioglossus anterior displacement and an increase in inferior displacement when the subjects inspired against a resistive load. This is similar to previous MRI studies observing more lateral narrowing of the upper airway and less anterior tongue motion during loaded inspiration (Cheng et al., 2011a; How et al., 2007). In previous studies of genioglossus EMG, adding inspiratory resistance resulted in increased inspiratory time, tidal volume and decreased airflow (Badr et al., 1990; Hudgel et al., 1987; Iber et al., 1982). Inspiratory resistive loading likely caused greater negative epiglottic pressure (Horner et al., 1991a; Horner et al., 1991b; Mathew et al.,

106

1982; Wheatley et al., 1993a), resulting in increased ventilatory drive to compensate with activation of upper airway dilator muscles and increased genioglossus EMG. These effects have been shown in wakefulness and sleep (Badr et al., 1990; Hudgel et al., 1987; Iber et al., 1982). This increased negative upper airway pressure appears to be a major driver of genioglossus activation. This is supported by findings that despite the increased intrapharyngeal pressure during resistive loading, the increase in pharyngeal resistance is only modest (Malhotra et al., 2002; Pillar et al., 2001). One hypothesis to explain the observed genioglossus motion is that moment to moment genioglossus activity serves to maintain upper airway patency, with overall motion determined by the balance of intraluminal negative pressure that acts to collapse it, and positive pressure generated by the upper airway dilatory muscles to open it (Remmers et al., 1978; Shepherd et al., 2006). The biomechanical effect of this dynamic motion would ensure airway patency is maintained within and between breaths (Bilston and Gandevia, 2014).

Movement Movement Genioglossus EMG (Ultrasound) (MRI)(Cheng et al., 2008, 2011a)

Spontaneous tidal  1 mm 0.5 – 2 mm Variable EMG breathing (anterior) (anterior) dependent of unit (tonic activation)

Voluntary hyperpnoea  Vt  ant  Vt mov’t  EMG activity

Inspiratory resistive load  resistance   resistance  ant resistance  EMG ant mov’t mov’t activity

Altered EELV  EELV  EELV

 ant mov’t  EMG activity

Table 4.8. Summary of genioglossus activity during different respiratory states. EMG – electromyography, Vt - tidal volume, mov’t – movement, EELV – end expiratory lung volume

During voluntary hyperpnoea genioglossus displacement increased during inspiration. We also observed a more homogeneous “en-bloc” antero-inferior movement during voluntary breathing, as compared to a heterogeneous pattern within the infero-posterior part of

107

genioglossus during spontaneous breaths. This is consistent with previous EMG studies that recorded higher genioglossus EMG in the posterior compared to anterior region in quiet breathing (Vranish and Bailey, 2015); greater genioglossus EMG without a significant difference between the most anterior and most posterior genioglossus regions during voluntary hyperpnoea or voluntary tasks when compared to quiet breathing; as well as greater differences in the phasic and tonic EMG components during voluntary tasks (Eastwood et al., 2003; Vranish and Bailey, 2015). The observed pattern of genioglossus motion may be the combined results of regional differences in muscle fiber type and composition between anterior and posterior genioglossus (Daugherty et al., 2012; Mu and Sanders, 2010; Saigusa et al., 2001), different and additional inputs from the motor cortex, and inputs from the pontine and parabrachial nuclei to the hypoglossal motor nucleus during voluntary tasks (Dutschmann and Dick, 2012; Laine and Bailey, 2011; Martelli et al., 2013; Sawczuk and Mosier, 2001).

Interestingly, when lung volume was altered by the external ventilator, a reduction or increase in end expiratory lung volume did not significantly change genioglossus displacement. This is in contrast to previous studies in which volitional or passive lowering of lung volumes during wakefulness occurred, whose findings include a reduction in upper airway size (Hoffstein et al., 1984; Series et al., 1990), increased genioglossus intramuscular EMG (Stanchina et al., 2003), and changes in tracheal traction and pharyngeal collapsibility (Rowley et al., 1996; Thut et al., 1993; Van de Graaff, 1988, 1991). Furthermore, we found a decrease in regional genioglossus displacement when six subjects were inside the external ventilator. One possible explanation may be the activation of chest wall mechanoreceptors receiving cutaneous and proprioceptive inputs by application of foam across the upper anterior chest wall to ensure adequate seal of the ventilation chamber (Davis, 1975; Nakayama et al., 1998).

In all our experiments where ventilatory drive was modulated, there were no significant differences between male and female subjects in the magnitude of maximal inspiratory infero-posterior regional genioglossus displacement. This is consistent with previous studies during wakefulness, although a significant increase in pharyngeal resistance (with or without inspiratory respiratory load) was observed in men during NREM sleep (Pillar et al., 2000a;

108

Trinder et al., 1997). In those studies, no gender differences were observed in central drive or response to loading, although greater genioglossus EMG was reported in awake women compared to men in one study (Popovic and White, 1995). Previous imaging studies have not observed significant sex-related differences in pharyngeal anatomy or pharyngeal dilator muscle activation during wakefulness in healthy subjects (Brooks and Strohl, 1992; Brown et al., 1986; Martin et al., 1997; White et al., 1985), or those with obstructive sleep apnoea (Rubinstein et al., 1989; Schwab et al., 1993).

4.1 Limitations Our study has limitations. First, the actions of the intrinsic and/or extrinsic tongue muscles may have contributed to the motion within the chosen infero-posterior part of genioglossus. In rodents, minimal intrinsic muscle activity is recorded during eupnoea (Bailey and Fregosi, 2004) but co-activation of intrinsic and extrinsic muscles can occur when drive is increased, as in hypoxia (Bailey et al., 2005). However, the region on which we focused shows the most motion during respiration (Cheng et al., 2008; Kwan et al., 2014). Second, epiglottic pressure, hypothesised to powerfully modulate central drive to genioglossus, was not measured in this study. However, our results are largely in keeping with previous MRI studies and genioglossus EMG studies, in which increased displacement occurs when ventilatory drive increases. Third, ultrasonography has intrinsic limitations. Detailed differentiation of the intramuscular architecture of genioglossus is difficult due to likely similar acoustic impedances within the muscle, and the image degradation produced by subject and transducer movements. However, we have previously demonstrated high intra- session and inter-session reliability with our method. Fourth, a variation in head and jaw position can influence the magnitude of measured posterior tongue motion during inspiration. In a recent MRI study, the largest motion was recorded with head in a neutral position standardised with the Frankfort plane and jaw open, and smallest with head extension. There was no significant difference between neutral and a mean flexion angle of 23, but there is a reduction of  0.5mm in measured motion at a mean extension angle of 23 from neutral position (Cai et al., 2016). In our study, the variation in head position is much smaller and likely to have minimal effect on measured motion. Finally, due to the paucity of data in utilising MRI or ultrasound to measure the magnitude of tongue movement in both healthy control and OSA subjects, a sample size of 20 subjects were chosen to participate in each

109

experiment. The results can now be added to data for prospective sample size calculations in future studies.

5. CONCLUSION Our findings provide improved understanding and quantification of genioglossus movement during conditions where ventilatory drive is altered in healthy subjects. Increased central drive did not always result in increased genioglossus displacement. The magnitude and direction of the infero-posterior regional genioglossus displacement is a result of the balance between the negative intraluminal pressure and the positive dilatatory pressure provided by the upper airway muscles in order to maintain upper airway patency. We have demonstrated the ability of ultrasound to record genioglossus movement in a research setting, with findings similar to previous MRI studies (Cheng et al., 2008, 2011a). Furthermore, ultrasonography can perform this in real time, at lower cost and greater accessibility, allow rapid assessment of the biomechanical effect of genioglossus activation, and allow comparison to previous EMG studies, thus providing more understanding of the complex dynamic system required to ensure upper airway patency.

110

Supporting Table 1. Mean maximal inspiratory displacement of 15 grid points during 6. SUPPLEMENTARYvoluntary hyperpnoe aTA expBLEerimen t for 20 subjects.

Point 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 A 0.49 0.51 0.50 0.52 0.55 0.60 0.62 0.58 0.54 0.60 0.93 0.95 0.91 0.72 0.62 B 0.47 0.48 0.44 0.36 0.36 0.58 0.49 0.45 0.39 0.36 0.54 0.46 0.42 0.37 0.37 C 0.21 0.18 0.23 0.23 0.19 0.21 0.13 0.21 0.22 0.24 0.15 0.28 0.28 0.32 0.28 0.30 0.25 0.22 0.27 0.21 0.22 0.34 0.33 0.34 0.35 0.19 0.38 0.33 0.33 0.36

g E 0.24 0.31 0.31 0.30 0.31 0.29 0.33 0.32 0.31 0.28 0.29 0.39 0.33 0.30 0.24 n

i F 0.76 0.78 0.63 0.52 0.49 0.98 0.94 0.74 0.60 0.57 1.15 1.28 1.00 0.97 0.85 h

t G 0.69 0.73 0.62 0.54 0.36 0.59 0.66 0.62 0.45 0.35 0.83 0.69 0.67 0.55 0.43 a

e H 0.64 1.29 1.34 1.31 1.22 0.61 1.21 1.44 1.28 1.25 0.66 1.13 1.54 1.65 1.38 r

b I 1.23 1.27 1.22 1.13 1.03 1.11 1.36 1.37 1.22 1.09 1.19 1.43 1.51 1.52 1.38

l J 0.20 0.19 0.15 0.13 0.14 0.18 0.21 0.18 0.19 0.16 0.21 0.30 0.28 0.27 0.26 a

d K 0.54 0.53 0.49 0.38 0.29 0.53 0.44 0.36 0.33 0.34 0.50 0.50 0.38 0.34 0.36

i t

L 0.72 0.72 0.70 0.67 0.71 0.78 0.82 0.81 0.88 0.88 0.92 1.01 1.00 0.96 0.73 s

u M 0.34 0.61 0.62 0.46 0.43 0.51 0.67 0.77 0.85 0.75 0.57 0.83 0.98 1.02 0.90 o 0.37 0.29 0.26 0.17 0.17 0.30 0.15 0.16 0.23 0.24 0.37 0.44 0.50 0.47 0.42 e N

n O 0.40 0.44 0.44 0.46 0.45 0.42 0.49 0.49 0.49 0.52 0.62 0.67 0.68 0.61 0.53 a

t P 0.87 0.89 0.84 0.83 0.76 0.98 1.00 0.98 0.99 1.01 1.05 1.10 1.13 1.14 1.25 n o Q 0.92 0.88 0.99 1.00 1.04 0.82 0.95 1.09 1.12 1.18 1.01 1.14 1.02 0.99 0.85

p R 0.92 0.95 0.95 0.88 0.78 0.86 0.89 0.92 0.84 0.81 0.67 0.86 0.86 0.77 0.67 S S 0.44 0.50 0.59 0.57 0.50 0.38 0.54 0.62 0.64 0.64 0.29 0.53 0.65 0.70 0.79 T 0.71 0.71 0.71 0.75 0.69 0.58 0.68 0.72 0.82 0.75 0.71 0.85 0.96 0.97 0.92 Mean 0.92 ± 1.01 ± 0.99 ± 0.93 ± 0.86 ± 0.93 ± 1.04 ± 1.06 ± 1.03 ± 1.00 ± 1.04 ± 1.23 ± 1.25 ± 1.21 ± 1.10 ± ± SD 0.45 0.52 0.53 0.52 0.50 0.44 0.55 0.60 0.56 0.54 0.53 0.56 0.62 0.65 0.58 A 0.70 0.77 0.74 0.81 0.69 0.26 0.52 0.58 0.55 0.58 0.46 0.27 0.33 0.29 0.40 B 0.20 0.21 0.22 0.24 0.20 0.19 0.19 0.19 0.17 0.13 0.26 0.27 0.27 0.21 0.17 C 0.76 0.76 0.61 0.46 0.27 0.44 0.64 0.71 0.60 0.43 0.21 0.40 0.36 0.35 0.36 D 0.03 0.04 0.02 0.06 0.08 0.09 0.10 0.10 0.13 0.17 0.10 0.11 0.07 0.04 0.03 0.09 0.13 0.13 0.22 0.22 0.12 0.16 0.16 0.14 0.10 0.17 0.22 0.13 0.16 0.17 E

g F 0.22 0.24 0.28 0.28 0.27 0.27 0.27 0.31 0.29 0.31 0.36 0.36 0.42 0.36 0.38 n

i G 0.98 0.91 0.80 0.63 0.40 1.04 0.97 0.78 0.58 0.35 0.89 0.82 0.73 0.55 0.39 h

t H 0.49 0.89 0.71 0.58 0.49 0.63 1.04 1.07 0.88 0.66 0.72 1.04 1.07 0.95 0.73 a

e I 1.01 0.81 0.79 0.73 0.80 1.02 1.00 0.90 0.93 0.93 0.93 1.22 1.25 1.20 1.21 r

b J 0.56 0.51 0.60 0.61 0.57 0.58 0.56 0.55 0.54 0.57 0.53 0.52 0.47 0.47 0.34 l 0.42 0.50 0.33 0.19 0.12 0.18 0.12 0.20 0.18 0.14 0.25 0.16 0.22 0.23 0.15 a K d 0.93 0.82 0.72 0.70 0.59 0.93 0.82 0.73 0.67 0.59 0.85 0.90 0.84 0.72 0.57

i L t

M 0.29 0.30 0.17 0.13 0.11 0.17 0.13 0.17 0.29 0.29 0.38 0.44 0.40 0.49 0.50 ”

t N 0.29 0.24 0.19 0.16 0.19 0.08 0.06 0.03 0.00 0.03 0.37 0.42 0.38 0.39 0.32 e

g O 0.35 0.35 0.42 0.38 0.34 0.42 0.36 0.42 0.31 0.21 0.33 0.30 0.22 0.15 0.21 r a P 0.54 0.53 0.47 0.40 0.38 0.55 0.50 0.45 0.39 0.39 0.64 0.58 0.54 0.48 0.39

T Q 1.01 0.98 0.90 0.75 0.76 1.02 1.09 1.00 0.80 0.77 1.14 1.24 0.98 0.60 0.39 “ R 0.14 0.15 0.22 0.24 0.19 0.21 0.20 0.22 0.19 0.16 0.20 0.22 0.27 0.29 0.31 S 0.92 1.22 1.38 1.41 1.31 0.72 1.11 1.23 1.26 1.37 0.51 0.89 1.00 1.12 1.19 T 1.11 1.27 1.31 1.21 1.08 1.12 1.28 1.33 1.27 1.16 1.08 1.28 1.28 1.22 1.19 Mean 0.89 ± 0.94 ± 0.89 ± 0.82 ± 0.73 ± 0.81 ± 0.90 ± 0.90 ± 0.82 ± 0.76 ± 0.84 ± 0.94 ± 0.91 ± 0.83 ± 0.76 ± ± SD 0.57 0.60 0.61 0.58 0.54 0.58 0.66 0.64 0.60 0.59 0.51 0.62 0.61 0.58 0.56

A 0.68 0.76 0.54 0.35 0.24 0.66 1.01 1.03 0.92 0.58 0.21 0.28 0.35 0.41 0.34

” t t 0.16 0.18 0.21 0.22 0.24 0.14 0.16 0.19 0.18 0.19 0.10 0.11 0.08 0.13 0.18

e g g 1.03 0.84 0.70 0.56 0.39 0.54 0.72 0.83 0.73 0.55 0.56 0.80 1.05 1.12 0.83

i a

a D 0.10 0.10 0.11 0.12 0.10 0.12 0.05 0.09 0.13 0.11 0.06 0.07 0.07 0.09 0.08

t i

“ E 0.35 0.31 0.24 0.25 0.27 0.30 0.31 0.27 0.26 0.27 0.33 0.29 0.49 0.43 0.28

a x

e 0.67 0.80 0.78 0.70 0.63 0.53 0.89 0.80 0.63 0.62 0.33 0.75 0.71 0.70 0.69

r 5

. G 0.65 0.49 0.27 0.07 0.10 0.55 0.45 0.29 0.25 0.10 0.54 0.45 0.37 0.28 0.20 b 1 H 1.94 2.01 1.80 1.61 1.11 1.88 2.23 1.96 1.79 1.17 1.57 2.21 1.96 1.70 1.33

Supplementary Table 4.S.1. Mean maximal inspiratory displacement of 15 grid points during voluntary hyperpnoea experiment for 20 subjects. Average maximal inspiratory displacement (mm) of 15 grid points located within genioglossus for 20 subjects. A – T denotes the 20 subjects. Data are expressed as mean ± SD.

111

Supporting Table 2. Mean maximal inspiratory displacement of 15 grid points during

inspiratory resistive load experiment for 20 subjects.

Point 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 A 0.90 0.96 0.61 0.64 0.49 0.85 1.02 0.96 0.80 0.44 0.95 1.19 1.08 0.65 0.51 B 0.46 0.70 0.76 0.69 0.49 0.44 0.59 0.69 0.49 0.63 0.86 0.80 0.94 0.87 0.63 C 1.08 1.14 1.12 1.05 0.87 1.15 1.23 1.24 1.03 1.02 1.39 1.42 1.36 1.13 1.04 1.11 1.23 0.99 0.92 0.93 1.12 1.24 1.26 1.24 1.24 0.60 0.93 1.22 1.22 0.98 g D

n E 0.65 0.80 0.68 0.72 0.54 0.59 0.63 0.64 0.59 0.85 0.42 0.54 0.56 0.47 0.47 i

h F 1.23 0.94 0.67 0.63 0.76 1.45 1.33 1.27 0.98 0.73 1.08 1.51 1.27 0.91 0.60 t

a G 0.91 0.87 0.78 0.80 0.65 1.08 1.14 0.96 1.02 0.92 0.88 1.33 1.35 1.38 1.26 e 0.83 0.80 0.70 0.76 0.83 0.93 0.91 0.83 0.95 0.88 1.03 1.06 1.10 1.07 0.98 r H

b I 1.39 1.18 0.92 0.84 0.68 1.56 1.69 1.49 1.25 0.78 1.31 1.74 1.69 1.10 0.49

l J 0.99 0.79 0.65 0.40 0.36 0.78 0.87 0.58 0.47 0.33 0.69 0.83 0.58 0.57 0.59 a

d K 0.56 0.78 0.92 0.92 0.70 0.57 0.83 0.89 0.68 0.48 0.46 0.60 0.58 0.50 0.37

i t

L 1.00 1.14 1.08 0.79 0.46 1.30 1.40 1.27 1.12 1.07 1.18 1.53 1.72 1.53 1.32

s M 0.61 0.76 0.63 0.47 0.50 0.39 0.75 0.99 0.68 1.06 0.51 0.62 0.40 0.26 0.99 u o N 2.10 2.94 2.64 1.89 1.34 2.61 3.20 2.94 2.37 1.87 2.16 3.23 2.96 1.89 1.28

e O 0.99 1.11 1.06 0.93 0.79 0.82 1.04 1.04 1.05 1.03 0.96 1.04 0.97 1.34 1.34 n

a P 0.72 0.72 0.64 0.66 0.61 0.70 0.87 0.98 0.95 0.87 0.61 1.13 1.24 1.17 1.06 t

n Q 0.88 0.79 0.76 0.64 0.52 0.90 0.99 0.74 0.57 0.50 0.92 0.89 0.69 0.38 0.22

o R 1.51 1.15 0.92 0.70 0.54 1.17 0.81 0.73 0.68 0.40 0.66 0.85 0.67 0.64 0.32 p

S S 0.21 0.24 0.29 0.31 0.28 0.30 0.34 0.39 0.41 0.26 0.29 0.33 0.32 0.22 0.21 T 0.43 0.49 0.44 0.27 0.26 0.43 0.57 0.31 0.46 0.39 0.18 0.23 0.31 0.43 0.26 Mean 0.93 ± 0.98 ± 0.86 ± 0.75 ± 0.63 ± 0.96 ± 1.07 ± 1.01 ± 0.89 ± 0.79 ± 0.86 ± 1.09 ± 1.05 ± 0.89 ± 0.75 ± ± SD 0.43 0.52 0.47 0.34 0.25 0.53 0.59 0.55 0.44 0.38 0.45 0.65 0.62 0.46 0.39 A 0.58 0.55 0.53 0.42 0.41 0.48 0.57 0.80 0.62 0.56 0.63 0.65 0.48 0.63 0.45 B 0.34 0.34 0.38 0.44 0.36 0.44 0.36 0.36 0.46 0.37 0.36 0.44 0.42 0.45 0.38 C 0.66 0.73 0.56 0.58 0.74 0.66 0.78 0.82 0.67 0.83 1.00 0.85 1.04 1.11 0.71 D 0.75 0.72 0.39 0.55 0.76 0.56 0.49 0.61 0.68 0.76 0.67 0.73 0.88 0.67 0.73 E 0.50 0.69 0.86 0.96 0.90 0.36 0.42 0.49 0.60 0.57 0.33 0.50 0.34 0.43 0.26 F 0.96 0.88 0.97 0.77 0.61 1.40 1.25 1.12 1.05 0.82 1.55 1.80 1.83 1.84 2.10 G 0.88 0.78 0.52 0.59 0.27 0.74 0.76 0.58 0.51 0.50 0.94 1.05 1.17 0.90 0.59 H 0.34 0.35 0.45 0.33 0.43 0.63 0.46 0.40 0.45 0.47 0.49 0.55 0.60 0.52 0.31

I 1.47 1.45 1.10 1.04 0.95 1.04 1.51 1.52 1.13 0.77 0.82 1.48 1.60 1.24 0.96

J 0.88 0.82 0.60 0.60 0.42 0.93 0.93 0.71 0.63 0.44 1.10 1.03 0.86 0.70 0.69 A

K 1.80 1.75 1.66 1.55 1.32 1.50 1.40 1.33 1.32 1.69 0.82 1.27 1.61 1.71 1.70 d a L 0.70 0.74 0.68 0.62 0.44 0.74 0.98 0.74 0.86 0.70 0.81 1.14 1.12 1.17 0.77 o M 0.39 0.33 0.22 0.20 0.18 0.38 0.45 0.40 0.28 0.30 0.12 0.38 0.43 0.46 0.24 L N 1.72 2.53 2.13 1.52 1.16 1.30 2.20 2.03 1.64 1.34 0.74 1.68 1.62 1.36 0.92 O 0.81 0.85 0.73 0.77 0.72 0.81 0.74 0.82 0.78 0.90 0.88 0.91 1.25 1.25 0.90 P 0.70 0.76 0.82 0.84 0.83 0.71 0.81 0.83 0.94 0.96 0.73 0.89 0.98 1.05 1.08 Q 0.55 0.52 0.42 0.41 0.46 0.48 0.65 0.70 0.70 0.70 0.51 0.67 0.45 0.49 0.37 R 2.44 2.01 1.72 1.30 0.87 1.91 1.79 1.50 0.84 0.59 1.02 1.07 1.04 0.72 0.39 S 0.59 0.61 0.73 0.74 0.65 0.39 0.64 0.67 0.70 0.70 0.43 0.43 0.46 0.61 0.62 T 0.22 0.28 0.35 0.44 0.40 0.24 0.38 0.36 0.39 0.34 0.21 0.42 0.49 0.38 0.26 Mean 0.86 ± 0.88 ± 0.79 ± 0.73 ± 0.64 ± 0.79 ± 0.88 ± 0.84 ± 0.76 ± 0.72 ± 0.71 ± 0.90 ± 0.93 ± 0.88 ± 0.72 ± ± SD 0.57 0.60 0.51 0.38 0.30 0.44 0.51 0.45 0.33 0.34 0.34 0.42 0.47 0.44 0.48 A 1.13 1.23 1.23 0.79 0.82 1.55 1.50 1.45 1.38 1.12 1.73 1.44 1.70 1.45 0.95 B 1.42 1.23 1.27 1.65 1.40 1.26 1.39 1.26 0.98 1.07 1.46 1.32 1.12 1.12 0.90 C 0.67 0.88 0.53 0.85 0.76 0.72 0.83 0.59 0.79 0.77 0.82 0.79 0.88 0.99 0.63 D 0.78 0.82 0.91 1.01 0.95 0.70 0.78 0.92 0.89 0.45 1.11 1.13 1.13 0.59 0.38

E 0.43 0.40 0.46 0.50 0.67 0.54 0.33 0.17 0.31 0.29 0.92 0.56 0.75 0.56 1.02 B

F 1.39 1.12 1.00 0.70 0.63 1.65 1.39 1.61 0.97 0.83 1.68 1.88 2.02 1.97 1.01

d G 0.60 0.46 0.24 0.07 0.05 0.48 0.40 0.43 0.18 0.23 0.64 0.59 0.59 0.52 0.34 a o H 0.50 0.43 0.49 0.46 0.52 0.48 0.50 0.51 0.56 0.61 0.43 0.48 0.62 0.67 0.55

L I 0.54 0.48 0.45 0.41 0.43 0.56 0.64 0.56 0.53 0.48 0.33 0.54 0.53 0.68 0.57 J 1.00 0.82 0.74 0.68 0.72 0.92 0.86 0.84 0.73 0.79 0.78 0.94 0.82 0.81 0.82 K 1.04 1.06 0.92 0.86 0.85 1.02 1.18 0.98 1.06 1.04 0.93 1.14 1.13 1.11 0.91 L 0.94 0.83 0.75 0.72 0.49 1.00 1.13 0.94 0.79 0.72 1.28 1.31 1.13 1.02 0.79 0.71 0.76 0.78 0.70 0.48 0.60 0.69 0.66 0.47 0.46 0.65 0.68 0.47 0.28 0.48 M Supplementary Table 4.S.2. Mean maximal inspiratory displacement of 15 grid points during inspiratory resistive load experiment for 20 subjects. Average maximal inspiratory displacement (mm) of 15 grid points located within genioglossus for 20 subjects. A – T denotes the 20 subjects. Load A, B and C added an inspiratory resistance of 11.6 cmH2O/L/s, 22.3 cmH2O/L/s, and 75.2 cmH2O/L/s respectively. Data are expressed as mean ± SD.

112

Supporting Table 3. Mean maximal inspiratory displacement of 15 grid points during

lung volume alteration experiment for 20 subjects.

Point 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 A 0.55 0.53 0.48 0.44 0.34 0.71 0.64 0.77 0.71 0.59 0.69 0.70 0.78 0.80 0.69

B 0.06 0.05 0.04 0.13 0.13 0.29 0.19 0.08 0.20 0.16 0.46 0.30 0.40 0.05 0.20

e C 0.13 0.19 0.10 0.07 0.14 0.21 0.21 0.18 0.06 0.12 0.12 0.17 0.17 0.09 0.08

r D 0.16 0.19 0.05 0.06 0.13 0.14 0.14 0.22 0.20 0.16 0.08 0.33 0.49 0.32 0.31 u

s E 0.33 0.32 0.44 0.32 0.24 0.36 0.30 0.43 0.31 0.48 0.38 0.42 0.48 0.54 0.35 s e F 2.04 1.80 1.02 0.64 0.58 1.23 1.02 0.65 0.49 0.42 1.18 1.22 0.89 0.59 0.37

r G 0.21 0.18 0.10 0.31 0.14 0.23 0.11 0.15 0.19 0.11 0.20 0.23 0.25 0.33 0.15

c H 0.11 0.22 0.29 0.24 0.15 0.22 0.07 0.19 0.24 0.17 0.27 0.23 0.16 0.26 0.27 i

c I 0.51 0.73 0.81 1.67 1.82 0.90 0.93 0.68 0.50 0.27 1.07 1.49 1.78 1.51 0.60

a J 0.11 0.00 0.05 0.07 0.20 0.09 0.04 0.04 0.09 0.11 0.10 0.10 0.14 0.03 0.07 r

o K 0.61 0.35 0.08 0.25 0.18 0.88 0.69 0.51 0.41 0.38 0.61 0.95 1.01 0.57 0.20 h

t L 0.19 0.19 0.26 0.24 0.14 0.09 0.39 0.22 0.28 0.15 0.02 0.42 0.40 0.28 0.16

a M 0.43 0.09 0.08 0.17 0.07 0.45 0.46 0.30 0.42 0.18 0.94 0.69 0.39 0.44 0.53 r

t N 0.16 0.11 0.17 0.44 0.18 0.10 0.08 0.45 0.29 0.20 0.63 0.45 0.56 0.48 0.17 x

e O 0.45 0.52 0.54 0.34 0.15 0.46 0.50 0.41 0.52 0.37 0.26 0.45 0.63 0.65 0.15

e P 0.06 0.08 0.04 0.02 0.27 0.63 0.34 0.17 0.32 0.34 0.46 0.53 0.30 0.24 0.05 v

i Q 0.22 0.34 0.22 0.23 0.26 0.39 0.42 0.41 0.42 0.30 0.73 0.74 0.61 0.47 0.24 t a R 0.26 0.56 0.50 0.19 0.11 0.42 0.69 0.70 0.50 0.08 0.42 0.76 1.03 0.98 0.83 g 0.13 0.11 0.10 0.08 0.07 0.13 0.12 0.10 0.09 0.08 0.09 0.09 0.09 0.10 0.11 e S

N T 0.23 0.18 0.15 0.08 0.14 0.31 0.27 0.23 0.11 0.41 0.39 0.51 0.11 0.05 0.09 Mean 0.35 ± 0.34 ± 0.28 ± 0.30 ± 0.27 ± 0.41 ± 0.38 ± 0.34 ± 0.32 ± 0.25 ± 0.46 ± 0.54 ± 0.53 ± 0.44 ± 0.28 ± ± SD 0.43 0.39 0.27 0.56 0.38 0.31 0.29 0.22 0.18 0.15 0.34 0.37 0.41 0.36 0.22 A 0.47 0.55 0.46 0.39 0.31 0.61 0.43 0.41 0.40 0.24 0.77 0.73 0.59 0.42 0.33 B 0.20 0.11 0.31 0.11 0.13 0.29 0.25 0.11 0.11 0.13 0.22 0.15 0.15 0.13 0.21

C 0.24 0.17 0.22 0.28 0.28 0.21 0.28 0.25 0.37 0.33 0.16 0.28 0.30 0.35 0.36

g D 0.45 0.58 0.48 0.30 0.23 0.53 0.46 0.50 0.44 0.30 0.41 0.73 0.62 0.48 0.25

n E 0.67 0.72 0.71 0.71 0.62 0.50 0.72 0.84 0.83 0.51 0.71 0.85 0.89 0.61 0.26 i

h F 0.87 0.63 0.50 0.50 0.37 1.20 1.05 0.61 0.55 0.69 1.19 1.43 1.10 0.47 0.26 t

a G 0.62 0.55 0.30 0.32 0.33 1.24 1.26 1.05 0.87 0.51 1.36 1.92 1.66 0.97 0.83 e

r H 0.46 0.35 0.40 0.22 0.18 0.48 0.43 0.41 0.39 0.36 0.45 0.58 0.48 0.48 0.55 b I

0.34 0.34 0.38 0.45 0.40 0.49 0.50 0.42 0.40 0.33 0.72 0.87 0.65 0.54 0.44

l J 0.12 0.09 0.14 0.11 0.08 0.11 0.10 0.09 0.03 0.11 0.11 0.23 0.19 0.20 0.28 a

d K 0.67 0.72 0.57 0.35 0.28 0.59 0.57 0.58 0.55 0.40 0.55 0.72 0.94 0.84 0.45

i t

L 0.22 0.20 0.30 0.21 0.23 0.22 0.32 0.30 0.36 0.33 0.29 0.18 0.28 0.33 0.26

s M 1.02 0.84 0.61 0.43 0.28 0.89 0.85 0.79 0.24 0.13 0.86 0.92 0.52 0.34 0.38 u o N 0.31 0.59 0.36 0.36 0.47 0.50 0.35 0.57 0.42 0.40 0.48 0.69 0.66 0.44 0.35

e O 0.22 0.17 0.14 0.24 0.26 0.36 0.31 0.21 0.14 0.10 0.24 0.47 0.49 0.39 0.25 n

a P 0.52 0.40 0.29 0.13 0.12 0.52 0.57 0.54 0.41 0.33 0.55 0.70 0.66 0.51 0.47 t

n Q 0.25 0.23 0.16 0.21 0.16 0.24 0.19 0.18 0.20 0.11 0.40 0.48 0.32 0.08 0.19

o R 0.24 0.24 0.17 0.18 0.12 0.23 0.35 0.29 0.39 0.28 0.45 0.62 0.59 0.46 0.38 p

S S 0.55 0.52 0.49 0.47 0.48 0.60 0.60 0.62 0.62 0.63 0.56 0.68 0.75 0.78 0.75 T 0.45 0.34 0.34 0.34 0.15 0.67 0.57 0.33 0.17 0.24 0.65 0.68 0.39 0.24 0.23 Mean 0.44 ± 0.42 ± 0.37 ± 0.32 ± 0.27 ± 0.52 ± 0.51 ± 0.45 ± 0.39 ± 0.32 ± 0.56 ± 0.70 ± 0.61 ± 0.45 ± 0.37 ± ± SD 0.24 0.22 0.16 0.15 0.14 0.30 0.29 0.25 0.22 0.17 0.32 0.41 0.35 0.22 0.17 A 0.72 0.49 0.15 0.28 0.38 1.09 0.68 0.29 0.18 0.50 0.95 1.12 0.64 0.18 0.17 B 0.11 0.19 0.12 0.08 0.08 0.26 0.24 0.31 0.10 0.09 0.18 0.16 0.12 0.25 0.04

C 0.34 0.05 0.14 0.32 0.24 0.17 0.06 0.15 0.12 0.16 0.18 0.28 0.15 0.03 0.08 c

i D 0.75 0.65 0.44 0.51 0.50 0.67 0.87 0.67 0.44 0.35 1.00 1.07 0.85 0.78 0.32

e e

a E 0.58 0.63 0.45 0.55 0.45 0.67 0.49 0.46 0.39 0.34 0.48 0.41 0.34 0.41 0.38

r v

i F 1.06 0.85 0.61 0.51 0.72 1.18 0.98 0.52 0.34 0.22 1.14 1.20 0.62 0.23 0.26

s i

h G 0.83 0.97 0.79 0.65 0.50 1.86 1.52 1.24 1.24 0.76 1.59 2.36 2.07 1.05 0.95

t e

o 0.48 0.50 0.38 0.43 0.65 0.53 0.44 0.50 0.53 0.57 0.89 0.88 0.82 0.76 0.83

a H

r P

Ip 0.19 0.19 0.50 0.53 0.42 0.30 0.31 0.34 0.35 0.33 0.56 0.54 0.45 0.45 0.40 t

x J 0.15 0.32 0.58 0.61 0.18 0.41 0.28 0.23 0.18 0.28 0.60 0.73 0.68 0.38 0.30 e K 0.77 0.82 0.75 0.56 0.26 0.74 0.47 0.21 0.13 0.13 0.58 0.43 0.23 0.25 0.32 L 0.03 0.05 0.14 0.02 0.11 0.05 0.10 0.09 0.09 0.24 0.08 0.05 0.06 0.08 0.05 M 1.34 1.05 0.82 0.67 0.34 1.30 1.33 0.91 0.69 0.21 1.13 1.57 1.07 0.71 0.40 Supplementary Table 4.S.3. Mean maximal inspiratory displacement of 15 grid points during lung volume alteration experiment for 20 subjects. Average maximal inspiratory displacement (mm) of 15 grid points located within genioglossus for 20 subjects. A – T denotes the 20 subjects. Extrathoracic negative pressure increased end-expiratory lung volume (EELV) by ~1000 mL and extrathoracic positive pressure reduced EELV by ~500 mL respectively. Data are expressed as mean ± SD.

113

Supporting Table 4. Respiratory variables for the lung volume alteration experiment.

Spontaneous Negative Positive Outside Inside tidal extra- extra- chamber chamber breathing thoracic thoracic tidal tidal pressure pressure breathing breathing Number of 20 20 20 6 6 subjects (n) Tidal volume 375 ± 163 374 ± 135 365 ± 235 416 ± 134 439 ± 166 (mL) Respiratory rate 15.4 ± 2.7 16.5 ± 3.2 17.4 ± 3.7 13.3 ± 1.9 13.8 ± 1.8 (breaths / min) Respiratory cycle 4.10 ± 0.81 3.81 ± 3.62 ± 4.58 ± 4.43 ± length (sec) 1.02 0.86 0.72 0.60 Inspiratory time 1.26 ± 0.31 1.13 ± 1.13 ± 1.19 ± 0.97 ± (sec) 0.40 0.27 0.31 0.41

Supplementary Table 4.S.4. Respiratory variables for the lung volume alteration Data are expressed as mean ± SD. experiment. Data are expressed as mean ± SD.

114

CHAPTER 5

Tongue movement during respiration in obstructive sleep apnoea under ultrasound imaging

SUMMARY

The pathogenesis of obstructive sleep apnoea (OSA) is multifactorial, and is likely due to instability in the mechanisms controlling pharyngeal patency. In the upper airway, the tongue provides support and patency to the pharyngeal space, with the genioglossus being the largest tongue muscle and the largest upper airway dilator. The aim of this study is to assess the dynamic motion of the tongue muscle with ultrasonography in healthy and OSA subjects during quiet breathing, and to investigate factors that may influence this motion in OSA subjects. We measured genioglossus motion with ultrasound in 46 awake subjects with a range of apnoea-hypopnoea index (AHI) in supine position. Our results revealed a greater peak inspiratory motion in the posterior tongue (1.16 ± 0.70 mm) than in the anterior tongue (0.27 ± 0.15 mm) during quiet breathing, and non-uniform motion within the posterior tongue grid in healthy controls (p = 0.014), with a loss of this regional variability in OSA subjects (p > 0.5). We observed 3 different inspiratory motion patterns in the posterior tongue: “en bloc” anterior movement, anterior movement of the “oropharyngeal” region, and “minimal” movement pattern, with further breath-to-breath variability within and across subjects with similar AHI. Using a multinomial logistic regression analysis, there was an association between body mass index (BMI) and tongue-based angle with the pattern of tongue motion (probability of the model chi-square (43.998) p < 0.001), and an association between pharyngeal length and BMI with OSA severity (probability of the model chi-square (92.475) p < 0.001). The different patterns of motion observed may represent the delicate moment-to- moment coordination of upper airway dilator muscles activity and drive in order to maintain pharyngeal patency within and between breaths.

115

1. INTRODUCTION

Obstructive sleep apnoea (OSA) is the commonest sleep-related breathing disorder and is characterised by repeated partial or complete upper airway obstruction in the presence of breathing effort (Kapur et al., 2017), with concurrent oxygen desaturation (Safar et al., 1959). It can lead to reduced neurocognitive performance and increased comorbidities. The prevalence of OSA varies depending on the definition, but is estimated to be between 3 – 7 percent in the general population (Kim et al., 2004; Peppard et al., 2013; Young et al., 2009; Young et al., 1993). The exact pathogenesis of OSA is unclear, but is most likely due to disadvantageous interactions between upper airway anatomy and sleep-related changes in upper airway behaviours, causing instability in balance of upper airway patency (Dempsey et al., 2010; Isono et al., 1997). Flow through the pharyngeal airway is dependent on the opposing dilatory force (upper airway dilators and tracheal traction) and collapsing force (negative inspiratory transmural pressure and upper airway tissue pressure) (Heinzer and Series, 2011; Remmers et al., 1978). To maintain pharyngeal patency within and between breaths, delicate moment to moment coordination of upper airway muscle activity and drive is required (Bilston and Gandevia, 2014). Electromyography (EMG) studies have contributed to our understanding of cortical drive and upper airway behaviour during sleep/wake states in healthy and OSA subjects (Malhotra and White, 2002). However, dynamic upper airway muscle movement in response to the neuro-muscular input during sleep/wake states is poorly understood.

The human tongue muscle provides support and patency to the pharyngeal space, and is identified to have different activation pattern depending on tasks (Lowe, 1980), and can respond to airway pressure changes related to respiration (Cheng et al., 2008; Saboisky et al., 2006). The largest muscle in the tongue is the genioglossus, and it is also the largest upper airway dilator (Smith et al., 2005). Previous single motor unit genioglossus EMG studies recorded a complex interaction of phasic inspiratory, phasic expiratory, and tonic drives at the hypoglossal motor nucleus (Saboisky et al., 2006), as well as suggestive evidence for larger motor unit territories in the anterior and superficial regions of genioglossus compared with more posterior regions (Luu et al., 2017), raising possibilities of regional activation within the genioglossus. This drive can be modulated by hypercapnia (Lo et al., 2006; Nicholas et al., 2010; Saboisky et al., 2010), hypoxaemia (Onal et al., 1981) and negative pharyngeal airway pressure (Fogel et al., 2001; Loewen et al., 2011; Malhotra et al., 2002).

116

In OSA subjects, studies during wakefulness (i) increase genioglossus EMG activity in response to negative pressure pulse(Berry et al., 2003), (ii) a higher level of phasic and tonic upper airway muscle activity (Fogel et al., 2001), (iii) earlier discharge, (iv) increased level of firing and (v) longer duration of motor unit action potentials in particular groups of genioglossus inspiratory motor units during quiet breathing compared to control subjects (Saboisky et al., 2015; Saboisky et al., 2012). These studies showed evidence of chronic partial denervation and remodelling of the genioglossus motor units, possibly as a result of a chronic neuropathic insult secondary to OSA (Bilston and Gandevia, 2014; Butler and Gandevia, 2008; Saboisky et al., 2015). However, it is unclear as to the mechanical consequence to the genioglossus and tongue.

Tongue motion has previously been measured with tagged magnetic resonance imaging (tMRI) technique (Cheng et al., 2008) and ultrasonography using cross-correlation tracking algorithm (Chien et al., 2017; Kwan et al., 2014). There is good agreement and consistency in quantifying tongue movement in OSA and healthy control subjects during wakefulness between the two modalities (Kwan et al., 2017); but ultrasound also recorded higher variation in the peak posterior tongue displacement and larger inter-breath variability within each individual, possibly due to better temporal and spatial resolution. Compared to MRI, ultrasonography has further advantages including lower cost, lower noise, greater availability, quicker to perform, more sensitive to recording dynamic changes, and able to allow dynamic real-time recordings. Both techniques reported peak inspiratory movement within the posterior tongue in young normal subjects of ~1 mm anteriorly during quiet breathing (Cheng et al., 2008; Kwan et al., 2014). This anterior movement is reduced during a loaded inspiration (Cheng et al., 2011a; Kwan et al., 2018). Ultrasound also recorded non-uniform anterior motion within the posterior tongue grid under both conditions, with maximal motion recorded in the infero-posterior region. A more homogeneous anterior motion was recorded in the same region during voluntary hyperpnoea (Kwan et al., 2018). In healthy subjects, tMRI also recorded increased inspiratory motion of the genioglossus with increased age, increased body mass index (BMI), reduced nasopharyngeal luminal size, and steeper tongue- base angle (Cheng et al., 2014). No significant association was found with pharyngeal length. In a study involving OSA subjects and BMI matched control, Brown et al. described 4 patterns of posterior tongue movement, determined according to the mean peak displacement of the nasopharyngeal and oropharyngeal points within the posterior tongue

117

during quiet breathing: (1) uniform “en bloc” anterior displacement > 1 mm; (2) “bidirectional” displacement of the tongue where the superior nasopharyngeal point moved posteriorly > 1 mm and inferior oropharyngeal point moved anteriorly > 1 mm; (3) “minimal” (< 1 mm) displacement of both points; and (4) “oropharyngeal” - minimal displacement of the nasopharyngeal point with > 1 mm anterior displacement of the oropharyngeal point. Although variability was observed within and between studied subjects, severe OSA subjects with an apnoea-hypopnoea index (AHI) >50 recorded minimal displacement (Brown et al., 2013a).

The aim of this study is to further understand the dynamic motion of the tongue muscle in healthy and OSA subjects during quiet breathing, and investigate potential factors that may influence this response in OSA subjects. The hypothesis was that (i) different patterns of movement would be recorded in OSA and healthy subjects within the posterior tongue, (ii) peak inspiratory motion would be recorded in the posterior tongue, with less anterior displacement in severe OSA subjects, (iii) regional variability in the magnitude of displacement within the posterior grid would be different between OSA and healthy controls and (iv) BMI, nasopharyngeal cross-sectional area, and tongue-base angle may influence the magnitude of posterior tongue motion during quiet breathing.

2. METHODS

Forty-six subjects with a range of AHI were recruited in this cross-sectional study. They were divided into 4 groups according to OSA severity: healthy control (AHI < 5), mild OSA (AHI 5 - 15), moderate OSA (AHI 15 - 30), and severe OSA (AHI ≥ 30) with 10 subjects in each group except 16 in moderate OSA group. All subjects had not received treatment for OSA prior to study. The groups were approximately matched by age, gender and BMI. Subjects with history of active or chronic respiratory or sleep disorder, upper airway surgery, and those using medication that could affect respiration, ventilation or upper airway muscles were excluded. All participants underwent a standard diagnostic polysomnogram in a clinical sleep laboratory. Diagnosis of OSA and determination of OSA severity were based on American Academy of Sleep Medicine guidelines (Berry et al., 2012). A 3T MRI scanner (Philips Achieva TX, Best, The Netherlands) was used to image the tongue and surrounding upper airway structures to provide anatomical measurements. Ethics approval was granted

118

by the Human Research Ethics Committees of the South East Sydney Area Health Service and University of New South Wales. The study was conducted according to the Declaration of Helsinki (2008) and informed written consent was obtained.

2.1 Experimental protocol

The experimental setup is similar to our previously published method (Kwan et al., 2014). Subjects lay supine in a standardised head and neck position with the Frankfort plane vertical to scanning table. Head, shoulder and neck padding was used to minimize effects on upper airway size (Amis et al., 1999; Trudo et al., 1998) and genioglossus motion (Cai et al., 2016) due to head flexion or extension. Subjects were instructed to stay awake, breathe through the nose, ensure mouth is closed and rest the tongue in its “neutral resting” position, usually with its tip on the incisors. To ensure wakefulness during the study, verbal communication with subjects occurred frequently.

Calibrated respiratory inductance bands (Inductotrace, Ambulatory Monitoring Inc., Ardsley, New York, USA) over thorax and abdomen were used to monitor the resting ventilation of each subject. Acquired signals were digitised at 1kHz (CED1401 data acquisition system and Spike2 software (Cambridge Electronic Design, Cambridge, UK)). B- mode ultrasound images were acquired in real-time at a frame rate of ~40 Hz with a Philips Healthcare iU22 system (Andover, USA) and a handheld curved array C8-5 transducer (frequency 5 to 8 MHz) positioned submentally in the mid-sagittal plane. Detailed description of the ultrasound protocol for imaging the respiratory cycle is published previously (Kwan et al., 2014). Three image “sequence” consisted of at least 5 consecutive stable consistent breaths were captured.

2.2 Image analysis

For tongue motion analysis, any sequence containing swallowing or jaw motion were excluded. Three consecutive breaths within the sequence with the clearest images were selected for offline analysis using a previously described custom developed 2D image cross- correlation MATLAB software (Mathworks, MA, USA) (Herbert et al., 2011; Kwan et al., 2014). Tongue motion was analysed with two rectangular grids (around 100mm2 area) consisted of 15 control “tracking points”, located in the anterior and posterior tongue, and one

119

other column positioned in the posterior tongue containing 5 control “tracking points” (≈ 5mm apart). Grid and column points nomenclature and location is given in Fig. 5.1. Point 1 of posterior-grid and posterior column was positioned ≈ 3 cm posterior to the internal spine of the mandible where there was peak inspiratory displacement reported in previous studies, and anterior-grid point 1 was positioned ≈ 2 cm superior to the internal spine of the mandible where there was minimal inspiratory displacement (Kwan et al., 2017; Kwan et al., 2014). A cross-correlation technique tracked the movement of these markers and provided the resultant inspiratory displacement of each “tracking point”, and overall displacement of grid / column. Inspiratory motion of the anterior and posterior groups within the posterior tongue grid was also assessed for variability.

Fig 5.1. Locations for tracking grids and posterior column and anatomical measurements. MRI and ultrasound images are of the same subject. Top left panel: ultrasound image of the placement of anterior and posterior grids, and posterior column. 15 solid red dots denote the control “tracking points” at start of inspiration for the grids. 5 solid blue dots denote the “tracking points” at start of inspiration for posterior column. Numerals denote the “tracking point” position within grid. Within grid: points 1-5 defined as “anterior” group, 6-10 as “middle” group, 11-15 as “posterior” group within grid. Points 11,12 defined as “infero-posterior” group. Within posterior column: points 1,2 defined as “inferior” group, 4,5 defined as “superior” group. Top middle panel: axial MRI image with blue line outlining the pharyngeal cross-sectional area at level of soft palate, through the narrowest part of the nasopharynx. Top right panel: sagittal MRI image denotes pharyngeal length and tongue-base angle measurements. Location used to measure pharyngeal cross-sectional area in axial view denotes by solid blue line.

120

Anatomical measurements of each subject’s tongue-base angle, pharyngeal airway length and luminal size were collected in the mid-sagittal and axial plane with 3T MRI (Achieva, Philips Medical Systems, Best, The Netherlands) at the narrowest point of the nasopharynx, similar to previously reported tMRI protocol (Brown et al., 2013a; Cheng et al., 2014; Cheng et al., 2008, 2011a). Patterns of posterior tongue movement during inspiration is categorised into four groups according to the description reported by Brown et al.(Brown et al., 2013a) The 4 patterns are determined according to the mean peak displacement of the superior (points 4 and 5) and inferior (points 1 and 2) groups within the posterior column across the 3 analysed breaths (Fig. 3): (1) uniform “en bloc” anterior displacement > 1 mm; (2) “bidirectional” displacement of the tongue where the superior group moved posteriorly > 1 mm and inferior group moved anteriorly > 1 mm; (3) “minimal” (< 1 mm) displacement of both groups; and (4) “oropharyngeal” - minimal displacement of the superior group with > 1 mm anterior displacement of the inferior group. The superior and inferior groups are similarly located to the previously described nasopharyngeal and oropharyngeal points under tMRI (Brown et al., 2013a; Cheng et al., 2008).

2.3 Statistical Analysis

Means and standard deviation (SD) were used for descriptive purposes. To assess differences in peak inspiratory displacement in different tongue regions and within posterior tongue grid across OSA severities, a one-way ANOVA with repeated measures with a Greenhouse-Geisser correction was used. Two-way ANOVA analysis was used to compare the effect of anatomical measurements on the magnitude of regional tongue movement. Post- hoc analyses were performed with Bonferroni’s correction, or Tukey’s test for pairwise comparison. To understand whether previously reported craniofacial and OSA risk factors that modulate genioglossus movement or pharyngeal airflow rate during inspiration (i.e. BMI, age, nasopharyngeal luminal size, tongue-base angle, pharyngeal length) can predict the posterior tongue movement pattern across healthy and OSA subjects, a multinomial logistic regression was performed. To evaluate any association between OSA severity and pattern of posterior tongue movement, a chi-square independence test was performed. Statistical significance was accepted at p < 0.05. All statistical analyses were carried out using SPSS version 24.0 (Armonk, NY, USA).

121

3. RESULTS

The mean age of the participants was 45.6 ± 14.2 years, with a mean BMI of 26.9 ± 6.1 kg/m2. Subject characteristics and anatomical measurements are summarised in Table 5.1. There was no significant difference between the groups in age, pharyngeal length, tongue- base angle and nasopharyngeal cross-sectional area. A significant difference was reported by one-way ANOVA analysis between the groups in BMI (F3,42 = 4.561, p = 0.007) and neck circumference (F3,42 = 6.584, p = 0.001). A Tukey post-hoc test revealed that the BMI was statistically lower in the healthy controls (24.9 ± 3.4 kg/m2, p = 0.012) and moderate severity OSA group (26.9.4 ± 4.6 kg/m2, p = 0.011) compared to the severe OSA group (32.9 ± 4.5 kg/m2); and a smaller neck circumference in health controls (36.5 ± 2.2 cm, p = 0.001) and mild severity OSA group (37.2 ± 4.7 cm, p = 0.005) compared to severe OSA group (42.6 ± 2.4 cm). Supplementary Tables are included after Conclusion.

Normal Mild Moderate Severe p value

(n=10) (n=10) (n=16) (n=10)

Age (years) 43.2 ± 15.9 41.1 ± 10.0 45.7 ± 13.7 55.9 ± 14.0 0.087

Gender (M : F) 6 : 4 6: 4 13 : 3 9: 1

AHI (events / h) 2.4 ± 1.7 11.0 ± 3.2 21.8 ± 4.1 58.4 ± 17.0

BMI (kg/m2) 24.9 ± 3.4 27.2 ± 5.1 26.9 ± 4.6 32.9 ± 4.5 0.007

Neck circumference 36.5 ± 2.2 37.2 ± 2.7 39.4 ± 4.7 42.6 ± 2.4 0.001 (cm)

Pharyngeal length 52.7 ± 5.6 53.4 ± 15.6 57.2 ± 9.0 64.7 ± 10.6 0.057 (mm)

Tongue-base angle (º) 70.5 ± 9.4 67.6 ± 9.8 71.1 ± 13.1 59.8 ± 14.9 0.130

Nasopharyngeal CSA 58.9 ± 28.2 46.6 ± 21.8 49.9 ± 56.6 43.6 ± 20.0 0.832 (mm2)

Table 5.1. Characteristics of the 46 subjects and their OSA severity. Body mass index (BMI). Apnoea-hypopnoea index (AHI). Cross-sectional area (CSA). Data are expressed as mean ± SD with the range given in brackets. See Methods for definition of OSA severity as defined by AHI. Separate one-way ANOVA analyses were performed to investigate any significant differences between the severity groups. Statistical significance was accepted at p < 0.05.

122

Table 5.2 summarises the location and characteristics of the grid. Tracking point 1 in relation to the internal mental spine of the mandible was on average ~20 mm for the anterior grid, ~30 mm for the posterior grid and ~33 mm for the posterior column from the reference point.

Distance (mm)

Posterior 15- Distance of point 1 from internal 27.1 ± 4.3 mental spine of the mandible point grid (mm)

Distance between grid columns 3.0 ± 0.5 (mm)

Distance between grid rows (mm) 2.1 ± 0.2

Posterior 5- Distance of point 1 from internal 31.3 ± 4.7 mental spine of the mandible point column (mm)

Distance between grid rows (mm) 5.5 ± 0.6

Anterior 15- Distance of point 1 from internal 21.7 ± 3.5 mental spine of the mandible point grid (mm)

Distance between grid columns 3.0 ± 0.3 (mm)

Distance between grid rows (mm) 2.0 ± 0.2

Table 5.2. Tongue grid / column characteristics. Data are expressed as mean ± SD. The internal mental spine of the mandible represents the insertion point of genioglossus.

3.1 Anterior grid

The mean peak inspiratory displacement across the 46 subjects measured 0.27 ± 0.15 mm, predominantly in the anterior direction. There was no significant difference observed across the OSA severity groups (F3,42 = 0.470, p = 0.705), and no significant regional variation in peak displacement within the grid (Table 5.3, Fig. 5.2).

123

Fig 5.2. Average maximal inspiratory movement of tracking grid in anterior and posterior tongue across healthy and OSA subjects. Solid black circles denote the 15 tracking points at the start of the respiratory cycle.

Mean maximal inspiratory displacement of tongue (mm)

AHI < 5 AHI 5 - 15 AHI 15 – 30 AHI ≥ 30

Anterior group 0.24 ± 0.16 0.22 ± 0.11 0.27 ± 0.17 0.32 ± 0.19

Middle group 0.27 ± 0.16 0.21 ± 0.13 0.29 ± 0.20 0.29 ± 0.17

Posterior group 0.26 ± 0.12 0.27 ± 0.17 0.30 ± 0.17 0.32 ± 0.19

Overall grid 0.25 ± 0.13 0.23 ± 0.12 0.28 ± 0.17 0.31 ± 0.18

p value 0.633 0.257 0.412 0.570

Table 5.3. Peak inspiratory tongue motion of different groups within the anterior grid.

Data are expressed as mean ± SD. See Methods for grid group definition. Overall grid described the mean of all 15 tracking points within grid. One-way ANOVA analysis performed to investigate any difference in peak inspiratory displacement of anterior, middle and posterior grid groups in healthy and OSA subjects. Statistical significance was accepted at p < 0.05.

3.2 Posterior grid

The posterior grid had a mean peak inspiratory displacement of 1.16 ± 0.70 mm across all subjects, predominantly in an anterior direction, with no significant difference between the

OSA severity groups (F3,42 = 1.167, p = 0.334) (Table 5.4). Non-uniform inspiratory motion was observed within the grid in normal healthy controls, with higher peak inspiratory

124

displacement of the posterior and middle groups compared with the anterior group (F2,1.535 = 6.522, p = 0.014). A more uniform anterior motion was observed across all three groups within the grid in OSA subjects with mild (p = 0.549), moderate (p = 0.528) and severe (p = 0.554) severity (Table 5.4, Fig. 5.2).

Mean maximal inspiratory displacement of tongue (mm)

AHI < 5 AHI 5 - 15 AHI 15 – 30 AHI ≥ 30

Anterior group 1.08 ± 0.75 0.83 ± 0.55 1.31 ± 0.84 1.35 ± 0.78

Middle group 1.18 ± 0.80 0.77 ± 0.41 1.33 ± 0.79 1.30 ± 0.59

Posterior group 1.25 ± 0.79 0.85 ± 0.45 1.25 ± 0.80 1.26 ± 0.60

Overall grid 1.17 ± 0.77 0.81 ± 0.45 1.29 ± 0.79 1.29 ± 0.65

p value 0.014 0.549 0.528 0.554

Table 5.4. Peak inspiratory tongue motion of different groups within the posterior grid.

3.3 Posterior column and tongue movement pattern

Across the 46 subjects, 138 image sequences of tongue motion were analysed. Fourteen of the 46 subjects exhibited a different inspiratory tongue motion pattern across the 3 imaged breath sequences. We found an association between OSA severity and posterior tongue movement pattern using chi-square independence test, χ2(6) = 15.38, p = 0.018 (Fig. 5.3, Supplementary Table 1). Overall, minimal movement (64/138, 46.4%) and oropharyngeal pattern (63/138, 45.7%) were the two predominant motion patterns (Fig. 5.3). The most common tongue movement pattern observed in mild and moderate OSA subjects is the “minimal” pattern (63.3% and 45.8% respectively), and in severe OSA subjects it is the “oropharyngeal” pattern (66.7%). Equal proportion of “minimal” and “oropharyngeal” patterns were observed in healthy controls (43.3%). The previously described bi-directional

125

pattern of movement was not observed (Brown et al., 2013a). Between breath variation in the tongue movement pattern was observed in 14 of the 46 subjects. Peak inspiratory movement of posterior column grid of four representative subjects is shown in Fig. 5.4.

Fig 5.3. Posterior tongue movement patterns and observed pattern distribution across different OSA severity.

Solid black circles denote the 5 tracking points at the start of the respiratory cycle. Each coloured line represents the mean displacement across 3 analysed breaths of an individual subject. The bar graph reports the observed tongue movement patterns across all 138 analysed image sequences for 46 subjects. Refer to Methods for a description of the movement patterns.

A multinomial logistic regression was performed to model the relationship between the OSA risk predictors and the 3 observed patterns of posterior tongue movement (“minimal”, “oropharyngeal” and “en bloc” anterior movement) across all analysed image sequences (Table 5.5). The traditional level of statistical significance 0.05 was used for all tests. The fit between the model and data was significantly improved by addition of the predictors to a model that only contained the intercept, 2 (10, n = 138) = 43.998, Nagelkerke R2 = 0.325, p < 0.001. Only tongue-base angle and BMI made significant unique contributions.

126

Fig 5.4. Posterior tongue movement patterns in 4 representative subjects. Solid black circles denote the 5 tracking points at the start of the respiratory cycle. Each coloured line represents the displacement across an analysed breath of an individual subject. Refer to Methods for a description of the movement patterns. Subjects A to C have AHI > 30. Subject D has AHI between 15 – 30. A. Two breaths have “minimal” pattern and one breath has “oropharyngeal” pattern. B. All three breaths have “oropharyngeal” pattern. C. All three breaths have “minimal” pattern. D. All three breaths have “en bloc” anterior movement pattern.

Input variable 2 df p

Pharyngeal length 178.486 2 0.799

Tongue-base angle 196.963 2 <0.001

Nasopharyngeal CSA 180.863 2 0.243

BMI 193.395 2 <0.001

Age 179.926 2 0.389

Table 5.5. Input variable’s unique contributions in the multinomial logistic regression (n = 138). CSA = cross-sectional area. 2 = amount by which -2 log likelihood increases when predictor is removed from the full model. Larger chi-square values indicate a poor fit for the model.

127

Tongue-base angle is the only statistically significant variable for comparing the “minimal” movement pattern with the “en bloc” anterior pattern, and BMI is the single significant parameter for comparing the “minimal” movement pattern with the “oropharyngeal” pattern (Table 5.6). The results suggest that (i) steeper tongue-base angle significantly reduce the likelihood of an “en bloc” anterior movement pattern in the posterior tongue, (ii) a higher BMI is significantly more likely to result in an “oropharyngeal” pattern compared to “minimal” movement pattern (odds ratio: OR = 0.85 and 1.12 respectively). En bloc anterior Oropharyngeal movement movement (n=11) (n=63)

Variable OR (95% CI) S.E. OR (95% CI) S.E.

Pharyngeal length 0.98 (0.89, 1.07) 0.05 0.99 (0.96, 1.03) 0.02

Tongue-base angle 0.85 (0.78, 0.93)* 0.05 0.97 (0.93, 1.01) 0.02

Nasopharyngeal CSA 1.02 (0.99, 1.04) 0.01 1.00 (0.99, 1.01) 0.01

BMI 0.88 (0.76, 1.01) 0.07 1.12 (1.03, 1.22)** 0.04

Age 0.98 (0.92, 1.03) 0.03 1.01 (0.98, 1.04) 0.02

Table 5.6. Multinomial logistic regression analysis for posterior tongue movement patterns and input variables. Reference group: “minimal” movement pattern (n=64). CSA = cross-sectional area. OR = odds ratio. S.E. = standard error. 95% CI = confidence interval. * p = 0.001, ** p = 0.011

Another multinomial logistic regression was performed to model the relationship between the OSA severity and (i) the 3 observed patterns of posterior tongue movement (“minimal”, “oropharyngeal” and “en bloc” anterior movement); (ii) OSA risk factors, across all analysed image sequences (Table 5.7). The fit between the model and data was significantly improved by addition of the input variables to a model that only contained the intercept, 2 (21, n = 138) = 92.475, Nagelkerke R2 = 0.523, p < 0.001. Age, pharyngeal length, BMI and type of posterior tongue movement pattern made significant unique contributions.

128

Input variable 2 df p

Pharyngeal length 301.627 3 <0.001

Tongue-base angle 290.211 3 0.085

Nasopharyngeal CSA 285.537 3 0.585

BMI 296.450 3 0.005

Age 294.920 3 0.010

Type of tongue movement pattern 300.554 6 0.009

Table 5.7. Input variable’s unique contributions in the multinomial logistic regression (n = 138).

CSA = cross-sectional area. 2 = amount by which -2 log likelihood increases when predictor is removed from the full model. Larger chi-square values indicate a poor fit for the model.

Pharyngeal length is the only statistically significant variable for comparing the healthy controls (AHI < 5) with moderate OSA (AHI between 5 to 15) and severe OSA (AHI > 30) subjects. BMI is the significant parameter for comparing the healthy controls with the severe OSA subjects (Table 5.8). The movement pattern was excluded as a variable because the 'en bloc' pattern occurred only in two OSA severity groups, causing quasi-separation in regression models. The results suggest that (i) longer pharyngeal length significantly increase the likelihood of moderate or severe OSA (OR = 1.06 and 1.14 respectively), (ii) a higher BMI is significantly more likely to result in severe OSA compared to healthy controls (AHI < 5) (OR = 1.28).

129

Mild OSA Moderate OSA Severe OSA Variable OR (95% CI) S.E. OR (95% S.E. OR (95% S.E. CI) CI) Pharyngeal 1.03 (0.96, 0.03 1.06 (1.00, 0.03 1.14 (1.06, 0.04 length 1.09) 1.13)* 1.23)**

Tongue-base 0.96 (0.90, 0.03 1.01 (0.97, 0.02 0.95 (0.88, 0.04 angle 1.02) 1.06) 1.02)

Nasopharyngeal 0.99 (0.98, 0.01 0.99 (0.98, 0.01 1.00 (0.98, 0.01 CSA 1.01) 1.00) 1.02)

BMI 1.10 (0.99, 0.06 1.04 (0.94, 0.05 1.28 (1.10, 0.08 1.23) 1.15) 1.50)***

Age 0.98 (0.92, 0.03 1.01 (0.98, 0.02 1.05 (0.99, 0.03 1.03) 1.04) 1.11)

Table 5.8. Multinomial logistic regression analysis for OSA severity and input variables.

Reference group: healthy controls AHI < 5 (n=10). CSA = cross-sectional area. OR = odds ratio. S.E. = standard error. 95% CI = confidence interval. * p = 0.048, ** p = 0.001, *** p = 0.002. See Methods for definition of OSA severity as defined by AHI.

4. DISCUSSION

This is the first ultrasound study to describe respiratory-related regional tongue motion patterns in OSA subjects during wakefulness. Our results recorded a greater peak inspiratory motion in the posterior tongue than in the anterior tongue during quiet breathing. We again recorded non-uniform motion within the posterior tongue grid in healthy controls, with larger displacements in the posterior group compared to anterior group (Kwan et al., 2018; Kwan et al., 2014). However, this regional variability was not observed in OSA subjects. Similar to the previous MRI study by Brown et al., we did find an association between the pattern of inspiratory-related tongue motion and different OSA severities, as well as breath-to-breath variability within subjects (Brown et al., 2013a). Furthermore, BMI and tongue-base angle may help predict the pattern of posterior tongue motion during inspiration. The different patterns of motion observed may represent the delicate moment-to-moment coordination of

130

upper airway dilator muscles activity and drive in order to maintain pharyngeal patency within and between breaths (Bilston and Gandevia, 2014).

4.1 Regional tongue displacement during inspiration

Consistent with previous MRI and ultrasound studies, maximal inspiratory motion of the tongue was recorded in the posterior region of around 1mm in an anterior direction, and significantly less in the anterior region (Brown et al., 2013a; Cheng et al., 2008; Kwan et al., 2018; Kwan et al., 2017; Kwan et al., 2014; Kwan et al., 2015). This was present in both healthy controls and severe OSA patients who have significantly higher BMI. The posterior tongue is largely made up of genioglossus muscle fibres, which interdigitates with intrinsic tongue muscles (Abd-El-Malek, 1939; Gaige et al., 2007; Saito and Itoh, 2007). The observed motion is consistent with the direction of the horizontal genioglossus fibers in this region, which pulls the tongue antero-inferiorly during an isolated contraction to dilate the upper airway (Smith et al., 2005). Increased BMI is also associated with increased genioglossus motion, a finding that confirms results from a prior MRI study (Cheng et al., 2014). An interesting observation however, is the loss of regional variation in the magnitude of inspiration motion within the posterior tongue grid in OSA subjects, irrespective of severity. The previously described non-uniform motion within the grid was recorded in young healthy control subjects during quiet breathing (Kwan et al., 2014), as well as during a loaded inspiration (Kwan et al., 2018). We believe the more uniform posterior tongue motion pattern observed in OSA subjects may be the combined result of different and additional inputs from the motor cortex (Dutschmann and Dick, 2012; Laine and Bailey, 2011; Martelli et al., 2013; Sawczuk and Mosier, 2001), as well as chronic neuropathic changes in genioglossus motor units, and altered tissue properties and airway mechanics. In OSA subjects, histological studies have recorded changes in upper airway muscle fibre types, distribution and composition (e.g. Boyd et al., 2004; Eckert et al., 2013; Edstrom et al., 1992; Woodson et al., 1991), as well as anatomical changes such as increased upper airway adiposity and soft tissue leading to increased collapsibility (Schwab et al., 2003; Stone et al., 2016). It may be related to the complex subcompartmental arrangement of motoneurones within the hypoglossal motor nucleus (Lowe, 1980), with the co-location of the motoneurones for the intrinsic tongue muscles and tongue protruders within the ventral compartment of the hypoglossal motor nucleus (Dobbins and Feldman, 1995; Gestreau et al.,

131

2005; McClung and Goldberg, 1999, 2002). This may ultimately act to increase lingual stiffness and tongue protrusion (e.g. Bailey et al., 2006; Doran, 1975; Stal et al., 2003) as a compensatory response of the upper airway dilator muscles to balance the increased collapsing forces present in OSA. However, a previous MR elastography study recorded lower tongue stiffness in awake OSA subjects, with no significant change during CPAP (Brown et al., 2015). In OSA there is likely some chronic neuropathic insult to genioglossus motor units, with subsequent remodelling of their fibres (e.g. Bilston and Gandevia, 2014; Butler and Gandevia, 2008; Saboisky, 2008; Saboisky et al., 2015; Saboisky et al., 2012). The observed motion pattern may also be caused by further changes in the upper airway muscle fibre types and properties as a compensatory adaptive response to increase demand to maintain pharyngeal patency in OSA (Boyd et al., 2004; Petrof et al., 1996; Smith et al., 2005; Woodson et al., 1991). Changes in the distribution of type II muscle fibres in association with reduced proportion of type I fibres has also been reported in genioglossus of OSA subjects, and is corrected by continuous positive airway pressure treatment (Carrera et al., 1999).

4.2 Effect of OSA risk factors on pattern of posterior tongue motion.

Obstructive sleep apnoea is a multifactorial disorder and most likely due to an imbalance between upper airway dilatory force and airway collapsing forces during sleep, with reduced upper airway patency as a consequence (Heinzer and Series, 2011). Risk factors for OSA include advancing age, male gender, obesity, and upper airway anatomical abnormalities (Peppard et al., 2013). Compared to matched control subjects, increased tongue volume and soft tissue structures within the upper airway were reported in OSA subjects (Kirkness et al., 2008; Schwab et al., 2003). Reduced pharyngeal cross-sectional area and shape also make the upper airway more collapsible (Burger et al., 1992a; Leiter, 1996; Rodenstein et al., 1990; Series et al., 1990). Previous tMRI studies in controls observed increased genioglossus displacement with increased age, increased body mass index (BMI), reduced nasopharyngeal luminal size, and sharper tongue-base angle (Cheng et al., 2014). Furthermore, OSA subjects with AHI > 50 were observed to commonly have minimal posterior tongue movement during inspiration, while age and BMI matched controls with AHI <5 predominantly have an “oropharyngeal” motion pattern (Brown et al., 2013a). In our study, we observed both “minimal” and “oropharyngeal” movement patterns in OSA subjects with AHI >30, and that

132

not all OSA risk factors are associated with the tongue motion. We also recorded breath-to- breath variability in the pattern of posterior tongue motion in 14 subjects, likely due to the delicate moment to moment coordination of upper airway muscles activity and drive to maintain pharyngeal patency (Bilston and Gandevia, 2014). In the present study, an increase in BMI is more likely to be associated with the “oropharyngeal” pattern of motion than “minimal”, and a subject with a steeper tongue-base angle is more likely to have a “minimal” motion pattern than “en bloc” anterior motion in the posterior tongue. It is likely both risk factors adversely affected the intricate balance of the upper airway patency. It is accepted that obesity is the strongest risk factor for development of OSA (Peppard et al., 2000a; Peppard et al., 2000b; Young et al., 2009). Prevalence of OSA is increased with increases in BMI and associated markers such as neck circumference and waist-to-hip ratio (Peppard et al., 2013; Peppard et al., 2000a; Young et al., 2004). Increased neck soft tissues surrounding the UA will likely increased transmural pressure and increase pharyngeal collapsibility (Schwab et al., 2003). Obesity also leads to increased adiposity in the tongue muscle and soft tissues surrounding the upper airway. Anatomical studies also observed a smaller airway and more tongue-base angulation in obese subjects (e.g. Isono, 2012; Mayer et al., 1996), possibly related to caudal displacement of the posterior tongue due to increase adiposity within the tongue. Pharyngeal length has also been revealed to have an association with OSA severity, with longer pharyngeal length recorded in more severe OSA subjects, and significantly more pharyngeal lengthening in OSA subjects after posture change from an upright to supine position. No significant change was revealed in subjects without OSA (Pae et al., 1997). The more frequent observation of “minimal” motion pattern in our subjects with more angled tongue-base may be due to a change in posterior tongue muscle fibre angulation, and in particular horizontal fibre fascicles. This may result in less anterior movement during contraction, and more cranio-caudal movement. However, no meaningful relationship between cranio-caudal or anterior movement and tongue-base angle was observed in this study. Furthermore, neural drive to inspiratory pump muscles has been revealed to be non-uniform (Saboisky et al., 2007c), and can be preferentially directed to muscles offering the greatest mechanical advantage to produce inspiratory airflow in a “neuromechanical matching” model (De Troyer et al., 2005; Gandevia et al., 2006).

The combination of increased BMI and increased tongue-base angle in our severe OSA subjects compares with healthy control may partly explain our findings of similar anterior

133

motion in the posterior region of the tongue, which is different from a previous MRI study which found predominantly “minimal” tongue motion in severe OSA subjects (Brown et al., 2013a). The observed differences can be due to anatomical and demographic differences in the studied subjects. For example, in our study, severe OSA subjects were older and had lower mean AHI then severe OSA subjects in the tMRI study. Our subjects’ mean BMI was also lower, and they had shorter pharyngeal length. This may also explain why a bidirectional tongue motion pattern was not observed in our study, which was previously hypothesised to be due to regional differences in mechanical load or an impaired coordination of posterior genioglossus muscle fibres due to non-uniform neuropathy (Brown et al., 2013a). Furthermore, our severe OSA subjects had a smaller nasopharyngeal airway size, flatter tongue-base angle, and significantly higher BMI compared to healthy control subjects, whereas the groups were more closely matched in the previous tMRI study. This would result in a need to actively dilate the upper airway during inspiration while awake and in a supine position in our severe OSA subjects. The effect of a less steep tongue-base may “improve” the biomechanical effectiveness of infero-posterior genioglossus muscle fascicles during a contraction, and increase the magnitude of tongue motion (Cheng et al., 2014). However, the presence of OSA would suggest that during sleep, this dilatory effect is unable to overcome the negative collapsing force. Potential causes may be reduced neuromuscular drive or dilatory muscles response (e.g. Jordan et al., 2007; Loewen et al., 2011; Mezzanotte et al., 1992; Remmers et al., 1978), lowered respiratory arousal threshold (e.g. Eckert et al., 2011b; Younes, 2004), and lowered activation threshold of the ventilatory control system (e.g. Wellman et al., 2004; Younes, 2008). Further research to investigate tongue motion of OSA subjects during sleep is needed to further our understanding of the functional response of upper airway muscles and whether this can help define an individual’s susceptibility to OSA and potential treatment options.

4.3 Limitations

In this ultrasound study, the major limitation of the technique is artifacts associated with ultrasonography. During submental imaging of the tongue, the hyoid and mandible can cause significant acoustic refraction. Intraluminal air within upper airway also results in difficulty to visualise structures deeper to the tongue surface. It is also important to stabilise the transducer as well as head position, as different head and jaw position can influence and

134

result in different magnitude of tongue motion during inspiration (Cai et al., 2016). Despite these limitations, our technique has good intra-rater repeatability within the same testing session as well as across sessions (Kwan et al., 2014), and has good agreement with tMRI in measuring tongue movement in subjects with and without OSA (Kwan et al., 2017).

Our findings also imply BMI and tongue-base angulation may be predictors to the pattern of posterior tongue inspiratory motion during wakefulness. These conclusions arise from a cohort with a significantly higher BMI in those with AHI  30, compared to those with AHI < 5 and AHI 15-30. Further investigation will be needed to identify other anatomical or biomechanical properties that influence the patency of the upper airway. We also recognise that the study was performed with subjects awake, and thus further imaging studies are required to investigate the motion of upper airway muscles during sleep.

4.4 Implications

Our findings contributed further to understanding the biomechanical properties of the tongue in healthy and OSA subjects during wakefulness. Previous studies have demonstrated increased or unchanged neural drive in OSA subjects during wakefulness (Malhotra et al., 2002; Saboisky et al., 2015; Saboisky et al., 2007a), but how that affects tongue motion is unclear. The advantages of ultrasound as a research imaging modality includes lower cost, lower noise, greater availability, quicker to perform, more sensitive to recording dynamic changes, and able to allow dynamic real-time recordings. It also has higher temporal and spatial resolution compared with MRI. By applying our ultrasound technique in this study, we recorded similar posterior tongue displacement in healthy and OSA subjects of  1mm, but intra-regional variability observed in healthy subjects were not recorded in OSA subjects. Furthermore, across the spectrum of OSA severities, the pattern of posterior tongue motion can vary both within a subject and across subjects with similar AHI, likely due to breath- breath balance of upper airway collapsing and dilatory forces to maintain patency. We have been able to identify BMI and tongue-base angulation as likely predictor of posterior tongue movement pattern during wakefulness, but further studies will be needed to identify pattern and magnitude of motion during sleep in healthy and OSA subjects, and to further our understanding of mechanisms involved in OSA pathogenesis.

135

5. SUPPLEMENTARY TABLE

Supplementary Table 5.S.1. Frequency and row percentages table reporting association between OSA severity and observed tongue movement pattern.

No OSA: AHI <5. Mild OSA: AHI 5 -15. Moderate OSA: AHI 15 – 30. Severe OSA: AHI  30. Refer to Methods for description of movement pattern.

136

CONCLUSION

137

CHAPTER 6

Summary of key findings and conclusion

1. GENERAL DISCUSSION

In this thesis, a novel ultrasound method to describe and quantify the biomechanical behaviour of the tongue in awake healthy and OSA subjects is presented. Real-time ultrasound recordings reveal variable movement within different tongue regions within and between subjects. This Chapter summarises the key findings which provide further understanding of the biomechanical function of the tongue, and highlights some future experiments to explore its dynamic behaviour during sleep / wake states in healthy and OSA subjects.

1.1 Dynamic tongue motion measured with ultrasonography

The ideal modality to image the upper airway should be easily accessible, inexpensive, noninvasive, free from radiation, able to be performed in the supine position during sleep, dynamic, and provide excellent image resolution (Schwab, 1998; Stuck and Maurer, 2008). However, all current image modalities have limitations and thus are not part of the standard diagnostic evaluation in usual clinical settings. MRI is currently considered the “best” upper airway imaging modality in OSA subjects but has disadvantages including high cost, noise, and limited availability (e.g. Fleetham, 1992; Kuo et al., 2011; Schwab, 1998; Togeiro et al., 2010). In comparison, ultrasound is also non-invasive and free from radiation, but it is easy to use, readily accessible, cheaper, and able to provide dynamic, high-resolution images (Stone, 2005). Hence, Chapters 2 and 3 investigated a novel ultrasonography method to image and quantify the tongue and genioglossus motion compared to MRI. The method has good intra-rater reliability and repeatability, and provides good agreement and consistency with an existing tagged-MRI method. The other advantage of ultrasound is its high spatial and temporal resolution, which allows the determination of motion within different regions of the tongue, and the timing of the posterior tongue motion relative to inspiratory airflow. We also showed inspiratory motion is maximal in the infero-posterior region of the tongue, with minimal motion in the anterior region. This may be due to differences in the drive to, and

138

composition of muscle fibres types between the two regions. The posterior region predominantly is composed of the horizontal genioglossus muscle fibres with larger fibre diameter and predominantly type I fibres (e.g. Daugherty et al., 2012; Miller et al., 2002; Saigusa et al., 2001; Sanders and Mu, 2013; Stal et al., 2003; Stone et al., 2016). In contrast, the anterior tongue has a different muscle fibre matrix and interdigitation of extrinsic and intrinsic tongue muscles, with more vertical fibres and predominantly type II fibres which contains fast Myosin heavy chain content (e.g. Granberg et al., 2010; Sanders et al., 2013; Sokoloff et al., 2010; Stone et al., 2016). There is also an increased amount of intramuscular connective tissue and fat in the anterior tongue region compared to posterior region (Miller et al., 2002), and higher capillary density in the posterior tongue suggesting the region to be more fatigue resistant (e.g. Crow and Ship, 1996; Granberg et al., 2010; Zaidi et al., 2013).

The observed motion is also likely influenced by the respiratory drive to the tongue muscles. Within the hypoglossal motor nucleus, a complex subcompartmental arrangement of motoneurones exists with different activation patterns depending on tasks (for review, see Lowe, 1980). The ventral compartment contains motoneurones for the protruders and the dorsal compartment for the retractors (e.g. Dobbins and Feldman, 1995; Gestreau et al., 2005; McClung and Goldberg, 1999). Within the ventral compartment, genioglossus motoneurones exist in two separate pools, one for horizontal fibres and one for vertical fibres. This provides a neuroanatomical substrate for differential outputs to regional sets of genioglossus muscle fibres. Recent genioglossus EMG study found suggestive evidence for larger motor unit territories in the anterior region and inferior region (superior to geniohyoid muscle) of genioglossus compared with more posterior regions (Luu et al., 2017), in contrast to previous histopathological studies which found a gradual increase in the size of muscle fibres from anterior to posterior genioglossus. Furthermore, the study found predominantly expiratory related motor units in the anterior genioglossus region. This raises the possibility of regional activation within genioglossus but further studies will be needed to assess this definitively.

These reported regional differences in muscle fibres and neural input to the tongue most likely contribute to the heterogeneous EMG activity recorded in genioglossus (e.g. Eastwood et al., 2003; Saboisky et al., 2006; Saboisky et al., 2007a; Saboisky et al., 2007c), although more phasic activity is described in the anterior tongue at rest, compared to more tonic contractions in the posterior tongue. The effect likely contributes to the ability of the tongue

139

to perform a wide range of movements during various tasks such as respiration, vocalisation and swallowing (Altschuler et al., 1994; McClung and Goldberg, 2002; Mu and Sanders, 2010).

Our ultrasound method is further able to differentiate the uniformity of regional motion within the tongue and this was found particularly in the posterior tongue. Although there is intra- and inter-subject variability in the magnitude and pattern of inspiratory genioglossus motion, the recorded movement is mainly in an anterior direction and it begins before inspiratory flow, consistent with previously described phasic EMG in genioglossus (e.g. Saboisky et al., 2006; Saboisky et al., 2007c) and MRI (Cheng et al., 2008). Breath-to-breath effects of respiratory drive and mechanics on the upper airway musculature is possibly related to individual anatomical differences. This may contribute to the observed inter- and intra-subject variability in respiratory motion of genioglossus, with the likely aim to ensure pharyngeal patency (e.g. Bilston and Gandevia, 2014; Brown et al., 2013a). In Chapters 2 to 5, significantly more anterior inspiratory motion occurred in the posterior area within the posterior tongue grid compared with the anterior area in healthy subjects. This region of the tongue is mostly composed of the genioglossus muscle, with contribution from other intrinsic tongue muscles (e.g. Abd-El-Malek, 1939; Gaige et al., 2007; Sanders and Mu, 2013; Stone et al., 2016; Takemoto, 2001). As a group, this regional variability was not observed in OSA subjects, irrespective of severity. This is possibly as a result of the contribution of additional respiratory drives to the genioglossus and intrinsic muscles. It may be relevant that there is co-location of the motoneurones for the intrinsic tongue muscles within the centrolateral nucleus of the ventral compartment of the hypoglossal motor nucleus (McClung and Goldberg, 2002). This may ultimately act to increase lingual stiffness and tongue protrusion (e.g. Bailey et al., 2006; Doran, 1975; Stal et al., 2003) as a compensatory response of the upper airway dilator muscles to balance the increased collapsing forces present in OSA. However, a previous MR elastrography study recorded lower tongue stiffness in awake OSA subjects, with no significant change during CPAP (Brown et al., 2015), suggesting this is insufficient to normalise collapsibility in OSA subjects. A peripheral factor that may be relevant is the innate thixotropic property of human skeletal muscles (Hill, 1968; Lakie and Robson, 1988; Lakie et al., 1984; Proske et al., 1993). This refers to a disproportionately high resistance within muscle fibres to imposed small movement, which reduces to low resistance after a brief larger amplitude movement. It is markedly affected by the history of

140

local contractions and length changes (Proske et al., 1993). This thixotropic property of muscle may have other implications for understanding tongue motion but these have not been formally investigated. Future research into understanding the biomechanical properties of the tongue tissue during sleep will be important to evaluate the mechanics of lingual muscles in OSA subjects.

1.2 Influence of respiratory drive and mechanics on tongue motion

In humans, the tongue muscles have different activation patterns depending on the task (for review, see Lowe, 1980). In conditions where respiratory neural drive is increased, genioglossus EMG also increases. These include hypercapnic and hypoxic conditions (e.g. Jordan et al., 2002; Klawe and Tafil-Klawe, 2003; Mateika et al., 1999; Onal et al., 1981; Parisi et al., 1987; Patrick et al., 1982), during inspiratory flow-resistive loading (e.g. Malhotra et al., 2000a; Pillar et al., 2001), and during voluntary hyperventilation when awake (e.g. Eastwood et al., 2003; Vranish and Bailey, 2015). However, the effect of increased neural drive on the magnitude and direction of tongue motion is lacking. Mammalian tongues are thought to act as a “muscular hydrostat”, that is they have a constant-volume, with contraction in one region leading to deformation of another region of the tongue body, thus contributing to movement. The complex muscular architecture within the tongue allows it to move as a whole unit, or by selective activation of different muscle groups results in rotation or twist motion and change shape simultaneously in multiple dimensions through coordinated interaction between the intrinsic and extrinsic muscles, with a constant-volume constraint (e.g. Gilbert et al., 2007; Kier and Smith, 1985; Smith and Kier, 1989). Simultaneous activation of longitudinally and transversely orientated muscle fibres can result in stiffening of the tongue (e.g. Mu and Sanders, 2010; Sanders and Mu, 2013).

In Chapter 4, we explored the hypothesis that increased genioglossus inspiratory movement would occur in three selected physiological conditions in which inspiratory neural drive is increased. Inspiration against a resistive load increased posterior tongue motion, but at the highest inspiratory resistance, there was less anterior displacement and more inferior displacement of the posterior tongue. This was similarly observed in a MRI study (Cheng et al., 2011a; How et al., 2007), and may be due to increased drive to both genioglossus and intrinsic tongue muscles to overcome the increased negative pharyngeal pressure (e.g. Horner et al., 1991a; Horner et al., 1991b; Mathew et al., 1982; Wheatley et al., 1993a). This increased negative upper airway pressure appears to be a major driver of genioglossus

141

activation (Malhotra et al., 2002; Pillar et al., 2001). During voluntary breathing, we showed that an increase in voluntary respiratory drive resulted in increased motion and a loss of regional variability within the posterior tongue. This is likely a combined result of different and additional inputs to the hypoglossal motor nucleus during voluntary tasks, and subsequent activation of different intrinsic and extrinsic tongue motoneurones (e.g. Dutschmann and Dick, 2012; Laine and Bailey, 2011; Martelli et al., 2013; Sawczuk and Mosier, 2001). The observed motion is also supported by genioglossus EMG studies during voluntary tasks that recorded higher genioglossus EMG in the posterior compared to anterior region in quiet breathing and greater genioglossus EMG without a significant difference between the two regions during voluntary hyperpnoea or voluntary tasks; as well as greater differences in the phasic and tonic EMG components during voluntary tasks (e.g. Eastwood et al., 2003; Pittman and Bailey, 2009; Vranish and Bailey, 2015). Finally, although lowering of lung volume is associated with reduced pharyngeal airway size and increased genioglossus EMG (e.g. Hoffstein et al., 1984; Series et al., 1990; Stanchina et al., 2003), our results did not reveal a significant increase in genioglossus motion. This could be due to the added influence of additional cortical and reflex inputs to hypoglossal motoneurones secondary to activation of chest wall mechanoreceptors by the ventilation chamber which was used to change lung volume (Davis, 1975; Nakayama et al., 1998).

1.3 Implications of the pattern of tongue motion in OSA subjects

OSA is a common sleep disorder, associated with the presence of partial or complete upper airway closure in the presence of breathing effort (for review, see Kapur et al., 2017). It is a multifactorial disorder, likely due to unfavourable interactions between upper airway anatomy and sleep-related changes in upper airway behaviours, causing instability in the balance of upper airway patency (e.g. Dempsey et al., 2010; Isono et al., 1997). One proposed pathophysiological cause is an impaired neuromuscular drive or response of upper airway dilators during sleep (e.g. Jordan et al., 2007; Loewen et al., 2011; Mezzanotte et al., 1992; Remmers et al., 1978; for review, see Eckert et al., 2013). In vivo, neural drive to the muscles fibres results in contraction. This can result in an isometric, concentric or eccentric contraction (e.g. Herzog et al., 2015; Huxley, 1957; Huxley and Simmons, 1971). Our results in Chapter 5 suggested both processes may be present in OSA subjects during wakefulness, and may be determined by individual anatomical factors such as BMI and angulation of the tongue base. In awake OSA subjects with AHI  5, both a “minimal” pattern of inspiratory

142

motion (< 1mm) as well as an “oropharyngeal” anterior motion of  1mm was recorded within the posterior tongue. This suggests that either the UA dilator muscles have an impaired response to neural drive (“minimal motion” at commencement of inspiration to dilate the upper airway) or that the response of the dilators is inadequate to counterbalance the increased negative intraluminal pharyngeal airway pressure during sleep (contraction of the posterior tongue or whole tongue is present during wakefulness).

However, if there is an impaired upper airway muscle response to neural drive, it is unclear from current literature whether this is due to a myopathic process involving the muscle, or a neuropathic process involving the neural drive and pathway to the muscle or both. A number of studies reveal dysfunction in upper airway muscle structure and function in OSA subjects compared with controls (e.g. Blumen et al., 2004; Boyd et al., 2004; Carrera et al., 1999; Carrera et al., 2004; Friberg et al., 1998; Kimoff, 2007; Series et al., 1995), but others have reported little difference or conflicting results (e.g. Eckert et al., 2011a; Eckert et al., 2007; Mortimore et al., 2000). However, the concept that a motor neuropathy causes a lack of motor response / contraction in UA dilatory muscles of OSA subjects rather than an upper airway musculature myopathy is supported by previous single motor unit genioglossus EMG which recorded structural neurogenic changes consistent with chronic partial denervation of muscle fibres (Saboisky et al., 2015; Saboisky et al., 2007a), muscle fibre abnormalities reported in histopathological studies (e.g. Edstrom et al., 1992; Friberg et al., 1998; Lindman and Stal, 2002), with reported normal muscle functions (e.g. Eckert et al., 2007; Mortimore et al., 2000). The upper airway sensorimotor impairment may occur as a result of local pathophysiological effects of OSA such as repeated vibration, hypoxia and inflammation (e.g. Eckert et al., 2007; Hagander et al., 2009; Sunnergren et al., 2011), and may contribute to the progression of OSA (Eckert et al., 2011a; Larsson et al., 1992) or reduced sensitivity to airway occlusion during sleep (Kimoff et al., 2001; Nguyen et al., 2005). Recent EMG studies also revealed a dissociation between genioglossus and tongue muscles EMG and their mechanical response during anaesthesia and sleep in OSA subjects (Dotan et al., 2015; Dotan et al., 2013), possibly due to ineffective neural drive. Our observed patterns of posterior tongue motion during inspiration in awake OSA subjects may support an inadequate or ineffective neural drive to maintain upper airway patency.

143

However, the above concept of a neuropathic process cannot fully explain our results. Compared with previous tagged MRI results which recorded a predominantly “minimal” inspiratory motion pattern within the posterior tongue in OSA subjects with AHI >50 (Brown et al., 2013a), we recorded the patterns of “minimal” motion and “oropharyngeal” anterior motion during inspiration. In a “muscular hydrostat” such as the tongue, it is possible our 2D image plane may not fully able to record 3D movement pattern, importantly lateral movement especially with the complex interdigitation of extrinsic and intrinsic muscle fibres within the posterior tongue. However, it is unknown why in some OSA subjects (or in different breaths within the same subject) there is an anterior “dilatory motion” and in others the tongue muscles motion is minimal. One possible hypothesis is that muscle activity is influenced by local anatomy and breath-breath pharyngeal conditions. Previous studies have shown the tongue muscles have a different activation pattern depending on tasks (e.g. Lowe, 1980; O'Dwyer et al., 1981), with variation in dilator muscle function between subjects (Hrycyshyn and Basmajian, 1972; Spiro et al., 1994; Vitti and Basmajian, 1977). Changes in upper airway size, the amount of soft tissues surrounding the upper airway, the amount of adiposity within the tongue muscles, and tongue base angulation in OSA subjects may affect the biomechanical properties of posterior genioglossus muscle fascicles during a contraction and thus the magnitude of inspiratory motion (e.g. Isono, 2012; Mayer et al., 1996; Schwab et al., 2003; Stone et al., 2016). The observed patterns of motion may also be related to different OSA phenotypes with associated differences in demography. This is especially important in subjects with AHI between 5 – 30, where a high degree of heterogeneity in disease patterns can be observed (Joosten et al., 2012). Different phenotypes may be associated with different local mechanical load or anatomy, which will impact on the individual’s tongue motion to counteract the negative upper airway pressure.

Compared with healthy controls, a significantly more uniform motion pattern was also observed within the posterior genioglossus grid in the OSA subjects. This may be a result of the contribution of additional respiratory drive to the genioglossus and intrinsic muscles within this region as a compensatory response to balance the increased collapsing forces present in OSA (Bilston and Gandevia, 2014). However, it may also be due to neuropathic changes within the region produced by genioglossus motor unit remodelling (Saboisky et al., 2015; Saboisky et al., 2007a; Saboisky et al., 2012). This may explain the similar finding of a more uniform posterior grid motion pattern in a same group of healthy subjects during

144

voluntary breathing tasks compared to spontaneous quiet breathing (see Chapter 4). The task-dependent observed motion in healthy subjects is more likely due to local differences in neural drive and muscle anatomy rather than a neuropathic change in genioglossus and intrinsic muscles. However, the magnitude of an effect due to neuropathic changes is not currently known. It is also possible that in those OSA subjects with minimal inspiratory tongue motion, local neuropathic processes as part of disease progression may affect the ability of the muscle to dilate the upper airway irrespective of neural drive (e.g. Edstrom et al., 1992; Friberg et al., 1998; Lindman and Stal, 2002; Saboisky et al., 2007a).

1.4 Limitations of the ultrasound method and future directions

To further our understanding in the biomechanical properties of the upper airway during sleep / wake state and its response in different upper airway disorders such as OSA, it is essential that future studies examine the magnitude and pattern of tongue motion across these conditions with an “ideal” upper airway image modality in a cohort of OSA subjects with a broad range of AHI. This thesis proposed a new ultrasound imaging protocol which has some advantages to other upper airway imaging modalities. Ultrasonography allows dynamic real-time recordings and has high temporal and spatial resolution. It also costs less, has lower noise and vibration, has greater accessibility and is quicker to perform. It can be performed during wakefulness and sleep and can provide anatomical and dynamic measurement of upper airway muscular structures in OSA subjects (Lahav et al., 2009; Shu et al., 2013).

However, there are a number of limitations inherent to ultrasound physics including image artifacts (Kremkau and Taylor, 1986) and the properties of different commercial ultrasound machines. Within the tongue musculature, it is difficult to visualise and differentiate the detailed intramuscular architecture due to similar acoustic impedance differences (Kremkau, 2016a). Also, strong acoustic impedance differences at the air-tissue interfaces, such as the tongue surface, reflects majority of the sound waves and structures (e.g. palate, lateral pharyngeal airway walls) distal to the interface cannot be seen. In the sagittal plane, bony landmarks (e.g. mandible and hyoid bones) cause high attenuation of ultrasound waves and acoustic shadows that obscure part of the tongue image (Stone, 2005). Image quality can also vary between subjects due to tongue properties such as fat and water content which may change refraction of sound waves at internal tissue-tissue interfaces (Stal et al., 2003; Stone,

145

2005). In obese subjects, the main acoustic limiting factors are the depth of insonation required and attenuation of ultrasound energy by adipose tissue. Selecting the optimal ultrasound equipment setting for image acquisition is important to improve penetration as well as improving the signal-to-background noise ratio. Possible different pre- and post- processing filters and techniques that can be used include tissue harmonic imaging, compound imaging and speckle reduction filters. Furthermore, the type of transducer can also affect the quality and field of view. The ideal frequency range for the transducers is between 3 and 7 MHz for tongue imaging (Bressmann, 2008; Stone, 2005). Lower frequencies result in coarser, low-contrast images and higher frequencies provide better details and resolution. However, attenuation increases with higher frequency and limits the depth of the field of view. Finally, temporal resolution is improved by higher scan frame rate. Our method collects images at a scan frame rate of 70 Hz. However, the output video frame rate of the ultrasound machine is at 30 frames per second, historically designed for input into an analogue video system (Stone, 2005). This mismatch may be inadequate for fast tongue motion and reduce its real temporal resolution (Li et al., 2005).

In future, improved technical developments in transducer and image analysis algorithms may allow increased precision in analysing the motion, as well as grid position in relation to a bony reference point. Improvements in video frame rate of ultrasound machines to match scan frame rate (80-90 Hz) should further improve temporal resolution and interpretation of fast tongue motions during volitional and respiratory activities (Stone, 2005). A high frequency transducer with the ideal aperture size and placed between the mandible and hyoid may allow full view of the tongue in mid-sagittal plane. Unlike MRI, ultrasound produces less noise and vibration and can potentially be performed on asleep subjects. Application of ultrasound elastography can further provide qualitative and quantitative assessment of tissue stiffness and elastic properties of the tongue (e.g. Gao et al., 2016; Shiina et al., 2015; for review, see Gennisson et al., 2013; Sigrist et al., 2017). Concurrent EMG studies on different parts of the tongue during dynamic ultrasound imaging will also provide more precise understanding of the relationship between neural drive and local response of the muscle, and thus further our understanding of the pathogenesis of OSA. The same technique can also be applied to investigate the biomechanics of the tongue muscles with OSA treatments such as CPAP, surgery, mandibular advancement splints or non-invasive hypoglossal neural stimulation. A recent tMRI study recorded 3 patterns of posterior tongue motion with

146

mandibular advancement (Brown et al., 2013b), with low AHI associated with more en-bloc anterior movement and high AHI associated with minimal posterior tongue motion. Dynamic ultrasound imaging of tongue motion during hypoglossal nerve stimulation or CPAP usage will provide further information on the neuromechanical response of tongue muscle to different OSA treatment modalities, and potentially improve understanding on which phenotype may respond to treatment.

Finally, the tongue is a 3D structure and appeared to act as a “muscular hydrostat”. It may be important to perform volumetric calculations or acquire dynamic 3D tongue movement to further our understanding of the biomechanical properties of the tongue and its response to different tasks and mechanical loads. Only recently are 3D anatomical tongue models being developed (e.g. Sanders and Mu, 2013; Stone et al., 2016; Woo et al., 2015; Woo et al., 2012). Under ultrasound, an approximate measurement can be made from repeated 2D measurements across different planes of view (e.g. Flowers et al., 2005; Yang and Stone, 2002). Better measurements may be achievable in the future by adopting 4D imaging technique currently available in cardiac and foetal ultrasonography (e.g. Montagnat and Delingette, 2005; Yagel et al., 2007).

Overall, our ultrasound technique has the capacity to provide high resolution recordings of dynamic regional tongue motion in healthy and OSA patients. At present, it does not yet have an obvious role in OSA management. However, in a research setting, its potential future application includes assessment of the mechanical behaviour of the tongue in OSA patients during sleep and after weight loss. It may also be used to assess the effect of OSA treatments such as CPAP or mandibular advancement splints on regional tongue movement. Finally, the recording of dynamic respiratory tongue motion may be useful in other upper airway disorders such as swallowing disorders.

2. CONCLUSIONS

This thesis has provided further understanding of the biomechanical properties of the tongue in healthy and OSA subjects during wakefulness through a novel ultrasound technique. We have demonstrated the different patterns of posterior tongue motion in healthy and OSA subjects during wakefulness. Within the posterior tongue, the inspiratory motion is

147

predominantly anterior, in keeping with the direction of horizontal fibres of the genioglossus and its role as the largest upper airway dilator. The likely effect of this motion is to counterbalance the negative collapsing forces in the upper airway, and its variation between breaths and between individuals is possibly due to local anatomy and biomechanical properties. During sleep, this dilatory motion is unable to fully overcome the collapsing forces in OSA subjects. This analysis re-emphasised the need for future research to investigate the biomechanical behaviour of the tongue in OSA subjects during sleep and to assess concurrent measures of neural drive. This should further our understanding of the pathogenesis of OSA.

148

REFERENCES

Abbey, N.C., Block, A.J., Green, D., Mancuso, A., Hellard, D.W., 1989. Measurement of pharyngeal volume by digitized magnetic resonance imaging. Effect of nasal continuous positive airway pressure. Am J Respir Crit Care Med 140, 717-723.

Abd-El-Malek, S., 1938. A contribution to the study of the movements of the tongue in animals, with special reference to the cat. J Anat 73, 15-30 11.

Abd-El-Malek, S., 1939. Observations on the morphology of the human tongue. J Anat 73, 201-210 203.

Abrishami, A., Khajehdehi, A., Chung, F., 2010. A systematic review of screening questionnaires for obstructive sleep apnea. Can J Anaesth 57, 423-438.

Akan, H., Aksoz, T., Belet, U., Sesen, T., 2004. Dynamic upper airway soft-tissue and caliber changes in healthy subjects and snoring patients. Am J Neuroradiol 25, 1846-1850.

Aksoz, T., Akan, H., Celebi, M., Sakan, B.B., 2004. Does the oropharyngeal fat tissue influence the oropharyngeal airway in snorers? Dynamic CT study. Korean J Radiol 5, 102-106.

Aldes, L.D., 1990. Topographically organized projections from the nucleus subceruleus to the hypoglossal nucleus in the rat: a light and electron microscopic study with complementary axonal transport techniques. J Comp Neurol 302, 643-656.

Altschuler, S.M., Bao, X., Miselis, R.R., 1994. Dendritic architecture of hypoglossal motoneurons projecting to extrinsic tongue musculature in the rat. J Comp Neurol 342, 538-550.

Amis, A., Prochazka, A., Short, D., Trend, P.S., Ward, A., 1987. Relative displacements in muscle and tendon during human arm movements. J Physiol 389, 37-44.

Amis, T.C., O'Neill, N., Wheatley, J.R., 1999. Oral airway flow dynamics in healthy humans. J Physiol 515 ( Pt 1), 293-298.

Ancoli-Israel, S., Klauber, M.R., Stepnowsky, C., Estline, E., Chinn, A., Fell, R., 1995. Sleep-disordered breathing in African-American elderly. Am J Respir Crit Care Med 152, 1946-1949.

Anderson, B., 2002. Interaction of ultrasound with tissue, Echocardiology: The normal examination and echocardiographic measurements, 2nd ed. Cardiotext Pub, Brisbane, pp. 2-6.

Anderson, R.J., 1881. The morphology of the muscles of the tongue and pharynx. J Anat Physiol 15, 382-391.

Armstrong, J., Leigh, M., Walton, I., Zvyagin, A., Alexandrov, S., Schwer, S., Sampson, D., Hillman, D., Eastwood, P., 2003. In vivo size and shape measurement of the human upper airway using endoscopic longrange optical coherence tomography. Opt Express 11, 1817-1826.

Armstrong, J.J., Leigh, M.S., Sampson, D.D., Walsh, J.H., Hillman, D.R., Eastwood, P.R., 2006. Quantitative upper airway imaging with anatomic optical coherence tomography. Am J Respir Crit Care Med 173, 226-233.

Asakawa, D.S., Nayak, K.S., Blemker, S.S., Delp, S.L., Pauly, J.M., Nishimura, D.G., Gold, G.E., 2003. Real-time imaging of skeletal muscle velocity. J Magn Reson Imaging 18, 734-739.

149

Aston-Jones, G., Bloom, F.E., 1981. Activity of norepinephrine-containing locus coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle. J Neurosci 1, 876-886.

Avruch, L., Cooperberg, P.L., 1985. The ring-down artifact. J Ultrasound Med 4, 21-28.

Axel, L., Dougherty, L., 1989. MR imaging of motion with spatial modulation of magnetization. Radiology 171, 841-845.

Axel, L., Goncalves, R.C., Bloomgarden, D., 1992. Regional heart wall motion: two-dimensional analysis and functional imaging with MR imaging. Radiology 183, 745-750.

Ayappa, I., Rapoport, D.M., 2003. The upper airway in sleep: physiology of the pharynx. Sleep Med Rev 7, 9-33.

Bacon, W.H., Krieger, J., Turlot, J.C., Stierle, J.L., 1988. Craniofacial characteristics in patients with obstructive sleep apneas syndrome. Cleft Palate J 25, 374-378.

Bacon, W.H., Turlot, J.C., Krieger, J., Stierle, J.L., 1990. Cephalometric evaluation of pharyngeal obstructive factors in patients with sleep apneas syndrome. Angle Orthod 60, 115-122.

Badr, M.S., Skatrud, J.B., Dempsey, J.A., Begle, R.L., 1990. Effect of mechanical loading on expiratory and inspiratory muscle activity during NREM sleep. J Appl Physiol 68, 1195-1202.

Badr, M.S., Toiber, F., Skatrud, J.B., Dempsey, J., 1995. Pharyngeal narrowing/occlusion during central sleep apnea. J Appl Physiol 78, 1806-1815.

Bailey, E.F., Fregosi, R.F., 2004. Coordination of intrinsic and extrinsic tongue muscles during spontaneous breathing in the rat. J Appl Physiol 96, 440-449.

Bailey, E.F., Fridel, K.W., Rice, A.D., 2007. Sleep/wake firing patterns of human genioglossus motor units. J Neurophysiol 98, 3284-3291.

Bailey, E.F., Huang, Y.H., Fregosi, R.F., 2006. Anatomic consequences of intrinsic tongue muscle activation. J Appl Physiol 101, 1377-1385.

Bailey, E.F., Janssen, P.L., Fregosi, R.F., 2005. PO2-dependent changes in intrinsic and extrinsic tongue muscle activities in the rat. Am J Respir Crit Care Med 171, 1403-1407.

Barsoumian, R., Kuehn, D.P., Moon, J.B., Canady, J.W., 1998. An anatomic study of the tensor veli palatini and dilatator tubae muscles in relation to eustachian tube and velar function. Cleft Palate Craniofac J 35, 101-110.

Basford, J.R., Jenkyn, T.R., An, K.N., Ehman, R.L., Heers, G., Kaufman, K.R., 2002. Evaluation of healthy and diseased muscle with magnetic resonance elastography. Arch Phys Med Rehabil 83, 1530-1536.

Bayliss, D.A., Viana, F., Berger, A.J., 1992. Mechanisms underlying excitatory effects of thyrotropin- releasing hormone on rat hypoglossal motoneurons in vitro. J Neurophysiol 68, 1733-1745.

Benard, M.R., Becher, J.G., Harlaar, J., Huijing, P.A., Jaspers, R.T., 2009. Anatomical information is needed in ultrasound imaging of muscle to avoid potentially substantial errors in measurement of muscle geometry. Muscle Nerve 39, 652-665.

150

Bennett, G.A., Hutchinson, R.C., 1946. Experimental studies on the movements of the mammalian tongue; the protrusion mechanism of the tongue (dog). Anat Rec 94, 57-72.

Berry, R.B., Budhiraja, R., Gottlieb, D.J., Gozal, D., Iber, C., Kapur, V.K., Marcus, C.L., Mehra, R., Parthasarathy, S., Quan, S.F., Redline, S., Strohl, K.P., Davidson Ward, S.L., Tangredi, M.M., American Academy of Sleep, M., 2012. Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events. Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine. J Clin Sleep Med 8, 597-619.

Berry, R.B., White, D.P., Roper, J., Pillar, G., Fogel, R.B., Stanchina, M., Malhotra, A., 2003. Awake negative pressure reflex response of the genioglossus in OSA patients and normal subjects. J Appl Physiol 94, 1875-1882.

Berthon-Jones, M., Sullivan, C.E., 1982. Ventilatory and arousal responses to hypoxia in sleeping humans. Am J Respir Crit Care Med 125, 632-639.

Berthon-Jones, M., Sullivan, C.E., 1984. Ventilation and arousal responses to hypercapnia in normal sleeping humans. J Appl Physiol 57, 59-67.

Bhattacharyya, N., Blake, S.P., Fried, M.P., 2000. Assessment of the airway in obstructive sleep apnea syndrome with 3-dimensional airway computed tomography. Otolaryngol Head Neck Surg 123, 444- 449.

Bilston, L.E., Gandevia, S.C., 2014. Biomechanical properties of the human upper airway and their effect on its behavior during breathing and in obstructive sleep apnea. J Appl Physiol 116, 314-324.

Bland, J.M., Altman, D.G., 1986. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1, 307-310.

Blissett, A., Prinz, J.F., Wulfert, F., Taylor, A.J., Hort, J., 2007. Effect of bolus size on chewing, swallowing, oral soft tissue and tongue movement. J Oral Rehabil 34, 572-582.

Blumen, M.B., de La Sota, A.P., Quera-Salva, M.A., Frachet, B., Chabolle, F., Lofaso, F., 2004. Tongue mechanical characteristics and genioglossus muscle EMG in obstructive sleep apnoea patients. Respir Physiol Neurobiol 140, 155-164.

Bobbert, M.F., Huijing, P.A., van Ingen Schenau, G.J., 1986. A model of the human triceps surae muscle-tendon complex applied to jumping. J Biomech 19, 887-898.

Bolsterlee, B., D'Souza, A., Gandevia, S.C., Herbert, R.D., 2017. How does passive lengthening change the architecture of the human medial gastrocnemius muscle? J Appl Physiol 122, 727-738.

Bolsterlee, B., Gandevia, S.C., Herbert, R.D., 2016a. Effect of transducer orientation on errors in ultrasound image-based measurements of human medial gastrocnemius muscle fascicle length and pennation. PLoS One 11, e0157273.

Bolsterlee, B., Gandevia, S.C., Herbert, R.D., 2016b. Ultrasound imaging of the human medial gastrocnemius muscle: how to orient the transducer so that muscle fascicles lie in the image plane. J Biomech 49, 1002-1008.

Bonnefous, O., Pesque, P., 1986. Time domain formulation of pulse-Doppler ultrasound and blood velocity estimation by cross correlation. Ultrason Imaging 8, 73-85.

151

Borowiecki, B., Pollak, C.P., Weitzman, E.D., Rakoff, S., Imperato, J., 1978. Fibro-optic study of pharyngeal airway during sleep in patients with hypersomnia obstructive sleep-apnea syndrome. Laryngoscope 88, 1310-1313.

Bosma, J.F., Hepburn, L.G., Josell, S.D., Baker, K., 1990. Ultrasound demonstration of tongue motions during suckle feeding. Dev Med Child Neurol 32, 223-229.

Boutet, C., Abdirahman Mohamed Moussa, S., Celle, S., Laurent, B., Barthelemy, J.C., Barral, F.G., Roche, F., 2016. Supra-epiglottic upper airway volume in elderly patients with obstructive sleep apnea hypopnea syndrome. PLoS One 11, e0157720.

Bowers, E.J., Morgan, D.L., Proske, U., 2004. Damage to the human quadriceps muscle from eccentric exercise and the training effect. J Sports Sci 22, 1005-1014.

Boyd, J.H., Petrof, B.J., Hamid, Q., Fraser, R., Kimoff, R.J., 2004. Upper airway muscle inflammation and denervation changes in obstructive sleep apnea. Am J Respir Crit Care Med 170, 541-546.

Bradley, T.D., Brown, I.G., Grossman, R.F., Zamel, N., Martinez, D., Phillipson, E.A., Hoffstein, V., 1986. Pharyngeal size in snorers, nonsnorers, and patients with obstructive sleep apnea. N Engl J Med 315, 1327-1331.

Brennick, M.J., 2013. Understanding airway tissue mechanics is a step towards improving treatments in OSA. Sleep 36, 973-974.

Brennick, M.J., Delikatny, J., Pack, A.I., Pickup, S., Shinde, S., Zhu, J.X., Roscoe, I., Kim, D.Y., Buxbaum, L.U., Cater, J.R., Schwab, R.J., 2014. Tongue fat infiltration in obese versus lean Zucker rats. Sleep 37, 1095-1102, 1102A-1102C.

Brennick, M.J., Parisi, R.A., England, S.J., 1997. Influence of preload and afterload on genioglossus muscle length in awake goats. Am J Respir Crit Care Med 155, 2010-2017.

Brennick, M.J., Pickup, S., Cater, J.R., Kuna, S.T., 2006. Phasic respiratory pharyngeal mechanics by magnetic resonance imaging in lean and obese zucker rats. Am J Respir Crit Care Med 173, 1031- 1037.

Brennick, M.J., Pickup, S., Dougherty, L., Cater, J.R., Kuna, S.T., 2004. Pharyngeal airway wall mechanics using tagged magnetic resonance imaging during medial hypoglossal nerve stimulation in rats. J Physiol 561, 597-610.

Bressmann, T., 2008. Quantitative assessment of tongue shape and movement using ultrasound imaging, In: Colantoni, L., Steele, J. (Eds.), 3rd Conference on laboratory approaches to Spanish phonology. Cascadilla proceedings project, Somerville, MA, pp. 101-106.

Brooks, L.J., Strohl, K.P., 1992. Size and mechanical properties of the pharynx in healthy men and women. Am J Respir Crit Care Med 146, 1394-1397.

Brown, E.C., Cheng, S., McKenzie, D.K., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2013a. Respiratory movement of upper airway tissue in obstructive sleep apnea. Sleep 36, 1069-1076.

Brown, E.C., Cheng, S., McKenzie, D.K., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2013b. Tongue and lateral upper airway movement with mandibular advancement. Sleep 36, 397-404.

152

Brown, E.C., Cheng, S., McKenzie, D.K., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2015. Tongue stiffness is lower in patients with obstructive sleep apnea during wakefulness compared with matched control subjects. Sleep 38, 537-544.

Brown, E.C., Hudson, A.L., Butler, J.E., McKenzie, D.K., Bilston, L.E., Gandevia, S.C., 2011. Single motor unit recordings in human geniohyoid reveal minimal respiratory activity during quiet breathing. J Appl Physiol 110, 1054-1059.

Brown, I.B., McClean, P.A., Boucher, R., Zamel, N., Hoffstein, V., 1987. Changes in pharyngeal cross- sectional area with posture and application of continuous positive airway pressure in patients with obstructive sleep apnea. Am J Respir Crit Care Med 136, 628-632.

Brown, I.G., Bradley, T.D., Phillipson, E.A., Zamel, N., Hoffstein, V., 1985. Pharyngeal compliance in snoring subjects with and without obstructive sleep apnea. Am J Respir Crit Care Med 132, 211-215.

Brown, I.G., Zamel, N., Hoffstein, V., 1986. Pharyngeal cross-sectional area in normal men and women. J Appl Physiol 61, 890-895.

Buchaillard, S., Brix, M., Perrier, P., Payan, Y., 2007. Simulations of the consequences of tongue surgery on tongue mobility: implications for speech production in post-surgery conditions. Int J Med Robot 3, 252-261.

Bujang, M.A., Adnan, T.H., 2016. Requirements for minimum sample size for sensitivity and specificity analysis. J Clin Diagn Res 10, YE01-YE06.

Burger, C.D., Stanson, A.W., Daniels, B.K., Sheedy, P.F., 2nd, Shepard, J.W., Jr., 1992a. Fast-CT evaluation of the effect of lung volume on upper airway size and function in normal men. Am J Respir Crit Care Med 146, 335-339.

Burger, C.D., Stanson, A.W., Sheedy, P.F., 2nd, Daniels, B.K., Shepard, J.W., Jr., 1992b. Fast-computed tomography evaluation of age-related changes in upper airway structure and function in normal men. Am J Respir Crit Care Med 145, 846-852.

Bushberg, J.T., Seibert, J.A., 2002. Ultrasound, In: Bushberg, J.T. (Ed.), The essential physics of medical imaging. Lippincott Williams & Wilkins, New York, pp. 469-554.

Butler, J.E., Gandevia, S.C., 2008. The output from human inspiratory motoneurone pools. J Physiol 586, 1257-1264.

Butler, J.E., Hudson, A.L., Gandevia, S.C., 2014. The neural control of human inspiratory muscles. Prog Brain Res 209, 295-308.

Caballero, P., Alvarez-Sala, R., Garcia-Rio, F., Prados, C., Hernan, M.A., Villamor, J., Alvarez-Sala, J.L., 1998. CT in the evaluation of the upper airway in healthy subjects and in patients with obstructive sleep apnea syndrome. Chest 113, 111-116.

Cai, M., Brown, E.C., Hatt, A., Cheng, S., Bilston, L.E., 2016. Effect of head and jaw position on respiratory-related motion of the genioglossus. J Appl Physiol 120, 758-765.

Capilouto, G.J., Frederick, E.D., Challa, H., 2012. Measurement of infant tongue thickness using ultrasound: a technical note. J Clin Ultrasound 40, 364-367.

153

Carlson, D.M., Onal, E., Carley, D.W., Lopata, M., Basner, R.C., 1995. Palatal muscle electromyogram activity in obstructive sleep apnea. Am J Respir Crit Care Med 152, 1022-1027.

Carrera, M., Barbe, F., Sauleda, J., Tomas, M., Gomez, C., Agusti, A.G., 1999. Patients with obstructive sleep apnea exhibit genioglossus dysfunction that is normalized after treatment with continuous positive airway pressure. Am J Respir Crit Care Med 159, 1960-1966.

Carrera, M., Barbe, F., Sauleda, J., Tomas, M., Gomez, C., Santos, C., Agusti, A.G., 2004. Effects of obesity upon genioglossus structure and function in obstructive sleep apnoea. Eur Respir J 23, 425- 429.

Carrier, D.R., Gregersen, C.S., Silverton, N.A., 1998. Dynamic gearing in running dogs. J Exp Biol 201, 3185-3195.

Carskadon, M.A., Dement, W.C., 2001. Normal human sleep: an overview, Principles and practice of sleep medicine, 5th ed. Elselvier Saunders, St. Louis, pp. 3-15.

Casey, D.M., 1983. Palatopharyngeal anatomy and physiology. J Prosthet Dent 49, 371-378.

Chamberlin, N.L., Eikermann, M., Fassbender, P., White, D.P., Malhotra, A., 2007. Genioglossus premotoneurons and the negative pressure reflex in rats. J Physiol 579, 515-526.

Chen, E.J., Jenkins, W.K., O'Brien, W.D., Jr., 1995. Performance of ultrasonic speckle tracking in various tissues. J Acoust Soc Am 98, 1273-1278.

Chen, H., Aarab, G., de Ruiter, M.H., de Lange, J., Lobbezoo, F., van der Stelt, P.F., 2016. Three- dimensional imaging of the upper airway anatomy in obstructive sleep apnea: a systematic review. Sleep Med 21, 19-27.

Chen, J.W., Chang, C.H., Wang, S.J., Chang, Y.T., Huang, C.C., 2014. Submental ultrasound measurement of dynamic tongue base thickness in patients with obstructive sleep apnea. Ultrasound Med Biol 40, 2590-2598.

Cheng, C.F., Peng, C.L., Chiou, H.Y., Tsai, C.Y., 2002. Dentofacial morphology and tongue function during swallowing. Am J Orthod Dentofacial Orthop 122, 491-499.

Cheng, S., Brown, E.C., Hatt, A., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2014. Healthy humans with a narrow upper airway maintain patency during quiet breathing by dilating the airway during inspiration. J Physiol 592, 4763-4774.

Cheng, S., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2008. Movement of the tongue during normal breathing in awake healthy humans. J Physiol 586, 4283-4294.

Cheng, S., Butler, J.E., Gandevia, S.C., Bilston, L.E., 2011a. Movement of the human upper airway during inspiration with and without inspiratory resistive loading. J Appl Physiol 110, 69-75.

Cheng, S., Gandevia, S.C., Green, M., Sinkus, R., Bilston, L.E., 2011b. Viscoelastic properties of the tongue and soft palate using MR elastography. J Biomech 44, 450-454.

Chi, L., Comyn, F.L., Mitra, N., Reilly, M.P., Wan, F., Maislin, G., Chmiewski, L., Thorne-FitzGerald, M.D., Victor, U.N., Pack, A.I., Schwab, R.J., 2011. Identification of craniofacial risk factors for obstructive sleep apnoea using three-dimensional MRI. Eur Respir J 38, 348-358.

154

Chiang, Y.C., Lee, F.P., Peng, C.L., Lin, C.T., 2003. Measurement of tongue movement during vowels production with computer-assisted B-mode and M-mode ultrasonography. Otolaryngol Head Neck Surg 128, 805-814.

Chien, C.Y., Chen, J.W., Chang, C.H., Huang, C.C., 2017. Tracking dynamic tongue motion in ultrasound images for obstructive sleep apnea. Ultrasound Med Biol 43, 2791-2805.

Ciscar, M.A., Juan, G., Martinez, V., Ramon, M., Lloret, T., Minguez, J., Armengot, M., Marin, J., Basterra, J., 2001. Magnetic resonance imaging of the pharynx in OSA patients and healthy subjects. Eur Respir J 17, 79-86.

Colebatch, J.G., Adams, L., Murphy, K., Martin, A.J., Lammertsma, A.A., Tochon-Danguy, H.J., Clark, J.C., Friston, K.J., Guz, A., 1991. Regional cerebral blood flow during volitional breathing in man. J Physiol 443, 91-103.

Colrain, I.M., 2005. The K-complex: a 7-decade history. Sleep 28, 255-273.

Connel, D.C., Fregosi, R.F., 1993. Influence of nasal airflow and resistance on nasal dilator muscle activities during exercise. J Appl Physiol 74, 2529-2536.

Crow, H.C., Ship, J.A., 1996. Tongue strength and endurance in different aged individuals. J Gerontol A Biol Sci Med Sci 51, M247-250.

Cunningham, D.P., Basmajian, J.V., 1969. Electromyography of genioglossus and geniohyoid muscles during deglutition. Anat Rec 165, 401-409.

Daugherty, M., Luo, Q., Sokoloff, A.J., 2012. Myosin heavy chain composition of the human genioglossus muscle. J Speech Lang Hear Res 55, 609-625.

Davidson, L., 2006. Comparing tongue shapes from ultrasound imaging using smoothing spline analysis of variance. J Acoust Soc Am 120, 407-415.

Davidson, T.M., 2003. The Great Leap Forward: the anatomic basis for the acquisition of speech and obstructive sleep apnea. Sleep Med 4, 185-194.

Davis, J.N., 1975. The response to stretch of human intercostal muscle spindles studied in vitro. J Physiol 249, 561-579.

De Troyer, A., Kirkwood, P.A., Wilson, T.A., 2005. Respiratory action of the intercostal muscles. Physiol Rev 85, 717-756. deBerry-Borowiecki, B., Kukwa, A., Blanks, R.H., 1988. Cephalometric analysis for diagnosis and treatment of obstructive sleep apnea. Laryngoscope 98, 226-234.

Dedhia, R.C., Strollo, P.J., Soose, R.J., 2015. Upper airway stimulation for obstructive sleep apnea: past, present, and future. Sleep 38, 899-906.

DeGusta, D., Gilbert, W.H., Turner, S.P., 1999. Hypoglossal canal size and hominid speech. Proc Natl Acad Sci U S A 96, 1800-1804.

Dempsey, J.A., Smith, C.A., 2014. Pathophysiology of human ventilatory control. Eur Respir J 44, 495- 512.

155

Dempsey, J.A., Veasey, S.C., Morgan, B.J., O'Donnell, C.P., 2010. Pathophysiology of sleep apnea. Physiol Rev 90, 47-112.

Dickson, D.R., Maue, W., 1970. Human vocal anatomy. Thomas, Springfield, Ill.

Dilley, A., Greening, J., Lynn, B., Leary, R., Morris, V., 2001. The use of cross-correlation analysis between high-frequency ultrasound images to measure longitudinal median nerve movement. Ultrasound Med Biol 27, 1211-1218.

Do, K.L., Ferreyra, H., Healy, J.F., Davidson, T.M., 2000. Does tongue size differ between patients with and without sleep-disordered breathing? Laryngoscope 110, 1552-1555.

Dobbins, E.G., Feldman, J.L., 1994. Brainstem network controlling descending drive to phrenic motoneurons in rat. J Comp Neurol 347, 64-86.

Dobbins, E.G., Feldman, J.L., 1995. Differential innervation of protruder and retractor muscles of the tongue in rat. J Comp Neurol 357, 376-394.

Doghramji, K., Jabourian, Z.H., Pilla, M., Farole, A., Lindholm, R.N., 1995. Predictors of outcome for uvulopalatopharyngoplasty. Laryngoscope 105, 311-314.

Donnelly, L.F., Surdulescu, V., Chini, B.A., Casper, K.A., Poe, S.A., Amin, R.S., 2003. Upper airway motion depicted at cine MR imaging performed during sleep: comparison between young patients with and those without obstructive sleep apnea. Radiology 227, 239-245.

Doran, G.A., 1975. Review of the evolution and phylogeny of the mammalian tongue. Acta Anat 91, 118-129.

Doran, G.A., Baggett, H., 1972. The genioglossus muscle: a reassessment of its anatomy in some mammals, including man. Acta Anat 83, 403-410.

Dotan, Y., Pillar, G., Schwartz, A.R., Oliven, A., 2015. Asynchrony of lingual muscle recruitment during sleep in obstructive sleep apnea. J Appl Physiol 118, 1516-1524.

Dotan, Y., Pillar, G., Tov, N., Oliven, R., Steinfeld, U., Gaitini, L., Odeh, M., Schwartz, A.R., Oliven, A., 2013. Dissociation of electromyogram and mechanical response in sleep apnoea during propofol anaesthesia. Eur Respir J 41, 74-84.

Doty, R.W., Richmond, W.H., Storey, A.T., 1967. Effect of medullary lesions on coordination of deglutition. Exp Neurol 17, 91-106.

Douglas, N.J., Jan, M.A., Yildirim, N., Warren, P.M., Drummond, G.B., 1993. Effect of posture and breathing route on genioglossal electromyogram activity in normal subjects and in patients with the sleep apnea/hypopnea syndrome. Am J Respir Crit Care Med 148, 1341-1345.

Douglas, N.J., White, D.P., Weil, J.V., Pickett, C.K., Martin, R.J., Hudgel, D.W., Zwillich, C.W., 1982a. Hypoxic ventilatory response decreases during sleep in normal men. Am J Respir Crit Care Med 125, 286-289.

Douglas, N.J., White, D.P., Weil, J.V., Pickett, C.K., Zwillich, C.W., 1982b. Hypercapnic ventilatory response in sleeping adults. Am J Respir Crit Care Med 126, 758-762.

156

Drakonaki, E., 2012. Ultrasound elastography for imaging tendons and muscles. J Ultrason 12, 214- 225.

Drakonaki, E.E., Allen, G.M., Wilson, D.J., 2012. Ultrasound elastography for musculoskeletal applications. Br J Radiol 85, 1435-1445.

Du Brul, E.L., 1976. Biomechanics of speech sounds. Ann N Y Acad Sci 280, 631-642.

Duffin, J., 2004. Functional organization of respiratory neurones: a brief review of current questions and speculations. Exp Physiol 89, 517-529.

Dutschmann, M., Dick, T.E., 2012. Pontine mechanisms of respiratory control. Compr Physiol 2, 2443-2469.

Dziewas, R., Schilling, M., Engel, P., Boentert, M., Hor, H., Okegwo, A., Ludemann, P., Ringelstein, E.B., Young, P., 2007. Treatment for obstructive sleep apnoea: effect on peripheral nerve function. J Neurol Neurosurg Psychiatry 78, 295-297.

Eastwood, P.R., Allison, G.T., Shepherd, K.L., Szollosi, I., Hillman, D.R., 2003. Heterogeneous activity of the human genioglossus muscle assessed by multiple bipolar fine-wire electrodes. J Appl Physiol 94, 1849-1858.

Eckert, D.J., Lo, Y.L., Saboisky, J.P., Jordan, A.S., White, D.P., Malhotra, A., 2011a. Sensorimotor function of the upper-airway muscles and respiratory sensory processing in untreated obstructive sleep apnea. J Appl Physiol 111, 1644-1653.

Eckert, D.J., Owens, R.L., Kehlmann, G.B., Wellman, A., Rahangdale, S., Yim-Yeh, S., White, D.P., Malhotra, A., 2011b. Eszopiclone increases the respiratory arousal threshold and lowers the apnoea/hypopnoea index in obstructive sleep apnoea patients with a low arousal threshold. Clin Sci 120, 505-514.

Eckert, D.J., Saboisky, J.P., Jordan, A.S., Malhotra, A., 2007. Upper airway myopathy is not important in the pathophysiology of obstructive sleep apnea. J Clin Sleep Med 3, 570-573.

Eckert, D.J., White, D.P., Jordan, A.S., Malhotra, A., Wellman, A., 2013. Defining phenotypic causes of obstructive sleep apnea. Identification of novel therapeutic targets. Am J Respir Crit Care Med 188, 996-1004.

Edmonds, L.C., Daniels, B.K., Stanson, A.W., Sheedy, P.F., 3rd, Shepard, J.W., Jr., 1992. The effects of transcutaneous electrical stimulation during wakefulness and sleep in patients with obstructive sleep apnea. Am J Respir Crit Care Med 146, 1030-1036.

Edstrom, L., Larsson, H., Larsson, L., 1992. Neurogenic effects on the palatopharyngeal muscle in patients with obstructive sleep apnoea: a muscle biopsy study. J Neurol Neurosurg Psychiatry 55, 916-920.

Edwards, B.A., Wellman, A., Sands, S.A., Owens, R.L., Eckert, D.J., White, D.P., Malhotra, A., 2014. Obstructive sleep apnea in older adults is a distinctly different physiological phenotype. Sleep 37, 1227-1236.

Eikermann, M., Jordan, A.S., Chamberlin, N.L., Gautam, S., Wellman, A., Lo, Y.L., White, D.P., Malhotra, A., 2007. The influence of aging on pharyngeal collapsibility during sleep. Chest 131, 1702- 1709.

157

Eisele, D.W., Smith, P.L., Alam, D.S., Schwartz, A.R., 1997. Direct hypoglossal nerve stimulation in obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 123, 57-61.

Elias, R.M., Bradley, T.D., Kasai, T., Motwani, S.S., Chan, C.T., 2012. Rostral overnight fluid shift in end-stage renal disease: relationship with obstructive sleep apnea. Nephrol Dial Transplant 27, 1569- 1573.

Ell, S.R., Jolles, H., Galvin, J.R., 1986. Cine CT demonstration of nonfixed upper airway obstruction. Am J Roentgenol 146, 669-677.

Ell, S.R., Jolles, H., Keyes, W.D., Galvin, J.R., 1985. Cine CT technique for dynamic airway studies. Am J Roentgenol 145, 35-36.

Epstein, L.J., Kristo, D., Strollo, P.J., Jr., Friedman, N., Malhotra, A., Patil, S.P., Ramar, K., Rogers, R., Schwab, R.J., Weaver, E.M., Weinstein, M.D., Adult Obstructive Sleep Apnea Task Force of the American Academy of Sleep, M., 2009. Clinical guideline for the evaluation, management and long- term care of obstructive sleep apnea in adults. J Clin Sleep Med 5, 263-276.

Erskine, R.J., Murphy, P.J., Langton, J.A., Smith, G., 1993. Effect of age on the sensitivity of upper airway reflexes. Br J Anaesth 70, 574-575.

Everson, C.A., Toth, L.A., 2000. Systemic bacterial invasion induced by sleep deprivation. Am J Physiol Regul Integr Comp Physiol 278, R905-916.

Ezzat, A.E., El-Shenawy, H.M., 2015. Palatopharyngeus the missing palatal muscles: Anatomical and physiological review. Ann Maxillofac Surg 5, 226-228.

Feldman, J.L., Del Negro, C.A., Gray, P.A., 2013. Understanding the rhythm of breathing: so near, yet so far. Annu Rev Physiol 75, 423-452.

Feldman, J.L., Mitchell, G.S., Nattie, E.E., 2003. Breathing: rhythmicity, plasticity, chemosensitivity. Annu Rev Neurosci 26, 239-266.

Feldman, J.L., Smith, J.C., Ellenberger, H.H., Connelly, C.A., Liu, G.S., Greer, J.J., Lindsay, A.D., Otto, M.R., 1990. Neurogenesis of respiratory rhythm and pattern: emerging concepts. Am J Physiol 259, R879-886.

Fercher, A.F., Drexler, W., Hitzenberger, C.K., Lasser, T., 2003. Optical coherence tomography - principles and applications. Rep Prog Phys 66, 239.

Ferguson, K.A., Ono, T., Lowe, A.A., Ryan, C.F., Fleetham, J.A., 1995. The relationship between obesity and craniofacial structure in obstructive sleep apnea. Chest 108, 375-381.

Fitzgerald, M.J., Law, M.E., 1958. The peripheral connexions between the lingual and hypoglossal nerves. J Anat 92, 178-188.

Fleetham, J.A., 1992. Upper airway imaging in relation to obstructive sleep apnea. Clin Chest Med 13, 399-416.

Flowers, H., Bressmann, T., Carmichael, B., Heng, C.L., Wong, W., 2005. Reconstructing 3D tongue movement trajectories from multiplanar paced ultrasound scans. Can Acoust 33, 50-51.

158

Fogel, R.B., Malhotra, A., Pillar, G., Edwards, J.K., Beauregard, J., Shea, S.A., White, D.P., 2001. Genioglossal activation in patients with obstructive sleep apnea versus control subjects. Mechanisms of muscle control. Am J Respir Crit Care Med 164, 2025-2030.

Fouke, J.M., Strohl, K.P., 1987. Effect of position and lung volume on upper airway geometry. J Appl Physiol 63, 375-380.

Fredberg, J.J., Wohl, M.E., Glass, G.M., Dorkin, H.L., 1980. Airway area by acoustic reflections measured at the mouth. J Appl Physiol 48, 749-758.

Fregosi, R.F., 2008. Influence of tongue muscle contraction and dynamic airway pressure on velopharyngeal volume in the rat. J Appl Physiol 104, 682-693.

Fregosi, R.F., Fuller, D.D., 1997. Respiratory-related control of extrinsic tongue muscle activity. Respir Physiol 110, 295-306.

Fregosi, R.F., Ludlow, C.L., 2014. Activation of upper airway muscles during breathing and swallowing. J Appl Physiol 116, 291-301.

Friberg, D., Ansved, T., Borg, K., Carlsson-Nordlander, B., Larsson, H., Svanborg, E., 1998. Histological indications of a progressive snorers disease in an upper airway muscle. Am J Respir Crit Care Med 157, 586-593.

Fucile, S., Gisel, E., Schanler, R.J., Lau, C., 2009. A controlled-flow vacuum-free bottle system enhances preterm infants' nutritive sucking skills. Dysphagia 24, 145-151.

Fuller, D., Mateika, J.H., Fregosi, R.F., 1998. Co-activation of tongue protrudor and retractor muscles during chemoreceptor stimulation in the rat. J Physiol 507 ( Pt 1), 265-276.

Fuller, D.D., Williams, J.S., Janssen, P.L., Fregosi, R.F., 1999. Effect of co-activation of tongue protrudor and retractor muscles on tongue movements and pharyngeal airflow mechanics in the rat. J Physiol 519 Pt 2, 601-613.

Fuller, P.M., Gooley, J.J., Saper, C.B., 2006. Neurobiology of the sleep-wake cycle: sleep architecture, circadian regulation, and regulatory feedback. J Biol Rhythms 21, 482-493.

Fung, S.J., Yamuy, J., Xi, M.C., Engelhardt, J.K., Morales, F.R., Chase, M.H., 2000. Changes in electrophysiological properties of cat hypoglossal motoneurons during carbachol-induced motor inhibition. Brain Res 885, 262-272.

Furusawa, K., Yamaoka, M., Kogo, M., Matsuya, T., 1991. The innervation of the levator veli palatini muscle by the glossopharyngeal nerve. Brain Res Bull 26, 599-604.

Gabaldon, A.M., Nelson, F.E., Roberts, T.J., 2004. Mechanical function of two ankle extensors in wild turkeys: shifts from energy production to energy absorption during incline versus decline running. J Exp Biol 207, 2277-2288.

Gaige, T.A., Benner, T., Wang, R., Wedeen, V.J., Gilbert, R.J., 2007. Three dimensional myoarchitecture of the human tongue determined in vivo by diffusion tensor imaging with tractography. J Magn Reson Imaging 26, 654-661.

Galloway, G., Fedok, F.G., Wiegand, D.A., Towfighi, J., 1991. Sternohyoid muscle biopsy. A diagnostic technique in respiratory failure of neuromuscular origin. Chest 100, 540-543.

159

Galvin, J.R., Rooholamini, S.A., Stanford, W., 1989. Obstructive sleep apnea: diagnosis with ultrafast CT. Radiology 171, 775-778.

Gandevia, S.C., Hudson, A.L., Gorman, R.B., Butler, J.E., De Troyer, A., 2006. Spatial distribution of inspiratory drive to the parasternal intercostal muscles in humans. J Physiol 573, 263-275.

Gandevia, S.C., Rothwell, J.C., 1987. Activation of the human diaphragm from the motor cortex. J Physiol 384, 109-118.

Gao, Z., Nakao, S., Ishihara, S., Funami, T., Kohyama, J., 2016. A pilot study on ultrasound elastography for evaluation of mechanical characteristics and oral strategy of gels. Journal of texture studies 47, 152-160.

Garcia, D., 2010. Robust smoothing of gridded data in one and higher dimensions with missing values. Comput Stat Data Anal 54, 1167-1178.

Geddes, D.T., Kent, J.C., Mitoulas, L.R., Hartmann, P.E., 2008. Tongue movement and intra-oral vacuum in breastfeeding infants. Early Hum Dev 84, 471-477.

Gennisson, J.L., Deffieux, T., Fink, M., Tanter, M., 2013. Ultrasound elastography: principles and techniques. Diagn Interv Imaging 94, 487-495.

Gerard, J.M., Ohayon, J., Luboz, V., Perrier, P., Payan, Y., 2005. Non-linear elastic properties of the lingual and facial tissues assessed by indentation technique. Application to the biomechanics of speech production. Med Eng Phys 27, 884-892.

Gestreau, C., Dutschmann, M., Obled, S., Bianchi, A.L., 2005. Activation of XII motoneurons and premotor neurons during various oropharyngeal behaviors. Respir Physiol Neurobiol 147, 159-176.

Gilbert, R.J., Napadow, V.J., Gaige, T.A., Wedeen, V.J., 2007. Anatomical basis of lingual hydrostatic deformation. J Exp Biol 210, 4069-4082.

Gillis, G.B., Biewener, A.A., 2002. Effects of surface grade on proximal hindlimb muscle strain and activation during rat locomotion. J Appl Physiol 93, 1731-1743.

Gleadhill, I.C., Schwartz, A.R., Schubert, N., Wise, R.A., Permutt, S., Smith, P.L., 1991. Upper airway collapsibility in snorers and in patients with obstructive hypopnea and apnea. Am J Respir Crit Care Med 143, 1300-1303.

Goss, S.A., Johnston, R.L., Dunn, F., 1978. Comprehensive compilation of empirical ultrasonic properties of mammalian tissues. J Acoust Soc Am 64, 423-457.

Granberg, I., Lindell, B., Eriksson, P.O., Pedrosa-Domellof, F., Stal, P., 2010. Capillary supply in relation to myosin heavy chain fibre composition of human intrinsic tongue muscles. Cells Tissues Organs 192, 303-313.

Greenspan, R.J., Tononi, G., Cirelli, C., Shaw, P.J., 2001. Sleep and the fruit fly. Trends Neurosci 24, 142-145.

Gugger, M., Molloy, J., Gould, G.A., Whyte, K.F., Raab, G.M., Shapiro, C.M., Douglas, N.J., 1989. Ventilatory and arousal responses to added inspiratory resistance during sleep. Am J Respir Crit Care Med 140, 1301-1307.

160

Guimaraes, K.C., Drager, L.F., Genta, P.R., Marcondes, B.F., Lorenzi-Filho, G., 2009. Effects of oropharyngeal exercises on patients with moderate obstructive sleep apnea syndrome. Am J Respir Crit Care Med 179, 962-966.

Hagander, L., Harlid, R., Svanborg, E., 2009. Quantitative sensory testing in the oropharynx: a means of showing nervous lesions in patients with obstructive sleep apnea and snoring. Chest 136, 481-489.

Hamlet, S.L., 1980. Ultrasonic measurement of larynx height and vocal fold vibratory pattern. J Acoust Soc Am 68, 121-124.

Hamlet, S.L., Reid, J.M., 1972. Transmission of ultrasound through the larynx as a means of determining vocal-fold activity. IEEE Trans Biomed Eng 19, 34-37.

Haponik, E.F., Smith, P.L., Bohlman, M.E., Allen, R.P., Goldman, S.M., Bleecker, E.R., 1983. Computerized tomography in obstructive sleep apnea. Correlation of airway size with physiology during sleep and wakefulness. Am J Respir Crit Care Med 127, 221-226.

Hariri, L.P., Applegate, M.B., Mino-Kenudson, M., Mark, E.J., Bouma, B.E., Tearney, G.J., Suter, M.J., 2013a. Optical frequency domain imaging of ex vivo pulmonary resection specimens: obtaining one to one image to histopathology correlation. J Vis Exp.

Hariri, L.P., Applegate, M.B., Mino-Kenudson, M., Mark, E.J., Medoff, B.D., Luster, A.D., Bouma, B.E., Tearney, G.J., Suter, M.J., 2013b. Volumetric optical frequency domain imaging of pulmonary pathology with precise correlation to histopathology. Chest 143, 64-74.

Hartmann, N., Martrette, J.M., Strazielle, C., Westphal, A., 1999. Dystonia musculorum mutation and myosin heavy chain expression in skeletal and cardiac muscles. J Cell Biochem 74, 90-98.

Hartmann, N., Martrette, J.M., Westphal, A., 2001. Influence of the Lurcher mutation on myosin heavy chain expression in skeletal and cardiac muscles. J Cell Biochem Suppl Suppl 36, 222-231.

Hawkins, D., Hull, M.L., 1990. A method for determining lower extremity muscle-tendon lengths during flexion/extension movements. J Biomech 23, 487-494.

Haxhiu, M.A., van Lunteren, E., Mitra, J., Cherniack, N.S., Strohl, K.P., 1984. Comparison of the responses of the diaphragm and upper airway muscles to central stimulation of the sciatic nerve. Respir Physiol 58, 65-76.

Hedemark, L.L., Kronenberg, R.S., 1982. Ventilatory and heart rate responses to hypoxia and hypercapnia during sleep in adults. J Appl Physiol Respir 53, 307-312.

Hedrick, W.R., Hykes, D.L., Starchman, D.E., 2004. Basic ultrasound physics, In: Hedrick, W.R., Hykes, D.L., Starchman, D.E. (Eds.), Ultrasound physics and instrumentation, 4th ed. Mosby, Boston.

Heinzer, R.C., Series, F., 2011. Normal physiology of the upper and lower airways, Principles and practice of sleep medicine, 5th ed. Elsevier Saunders, St. Louis, Missouri, pp. 259-268.

Hellstrand, E., 1980. Morphological and histochemical properties of tongue muscles in cat. Acta Physiol Scand 110, 187-198.

Henry, L.R., Solomon, N.P., Howard, R., Gurevich-Uvena, J., Horst, L.B., Coppit, G., Orlikoff, R., Libutti, S.K., Shaha, A.R., Stojadinovic, A., 2008. The functional impact on voice of sternothyroid muscle division during thyroidectomy. Ann Surg Oncol 15, 2027-2033.

161

Herbert, R.D., Clarke, J., Kwah, L.K., Diong, J., Martin, J., Clarke, E.C., Bilston, L.E., Gandevia, S.C., 2011. In vivo passive mechanical behaviour of muscle fascicles and tendons in human gastrocnemius muscle-tendon units. J Physiol 589, 5257-5267.

Herbert, R.D., Gandevia, S.C., 1995. Changes in pennation with joint angle and muscle torque: in vivo measurements in human brachialis muscle. J Physiol 484 ( Pt 2), 523-532.

Herbert, R.D., Heroux, M.E., Diong, J., Bilston, L.E., Gandevia, S.C., Lichtwark, G.A., 2015. Changes in the length and three-dimensional orientation of muscle fascicles and aponeuroses with passive length changes in human gastrocnemius muscles. J Physiol 593, 441-455.

Herbert, R.D., Moseley, A.M., Butler, J.E., Gandevia, S.C., 2002. Change in length of relaxed muscle fascicles and tendons with knee and ankle movement in humans. J Physiol 539, 637-645.

Heroux, M.E., Stubbs, P.W., Herbert, R.D., 2016. Behavior of human gastrocnemius muscle fascicles during ramped submaximal isometric contractions. Physiol Rep 4.

Herzog, W., Powers, K., Johnston, K., Duvall, M., 2015. A new paradigm for muscle contraction. Front Physiol 6, 174.

Hida, W., Okabe, S., Miki, H., Kikuchi, Y., Taguchi, O., Takishima, T., Shirato, K., 1994. Effects of submental stimulation for several consecutive nights in patients with obstructive sleep apnoea. Thorax 49, 446-452Hilaire, G., Pasaro, R., 2003. Genesis and control of the respiratory rhythm in adult mammals. News Physiol Sci 18, 23-28.

Hill, D.K., 1968. Tension due to interaction between the sliding filaments in resting striated muscle. The effect of stimulation. J Physiol 199, 637-684.

Hochban, W., Brandenburg, U., Peter, J.H., 1994. Surgical treatment of obstructive sleep apnea by maxillomandibular advancement. Sleep 17, 624-629.

Hoffstein, V., Fredberg, J.J., 1991. The acoustic reflection technique for non-invasive assessment of upper airway area. Eur Respir J 4, 602-611.

Hoffstein, V., Zamel, N., Phillipson, E.A., 1984. Lung volume dependence of pharyngeal cross- sectional area in patients with obstructive sleep apnea. Am J Respir Crit Care Med 130, 175-178.

Holmer, N.G., Rundqvist, H.E., 1975. Ultrasonic registration of the fundamental frequency of a voice during normal speech. J Acoust Soc Am 58, 1073-1077.

Holstege, G., Kuypers, H.G., Dekker, J.J., 1977. The organization of the bulbar fibre connections to the trigeminal, facial and hypoglossal motor nuclei. II. An autoradiographic tracing study in cat. Brain 100, 264-286.

Horner, R.L., 1996. Motor control of the pharyngeal musculature and implications for the pathogenesis of obstructive sleep apnea. Sleep 19, 827-853.

Horner, R.L., 2001. The neuropharmacology of upper airway motor control in the awake and asleep states: implications for obstructive sleep apnoea. Respir Res 2, 286-294.

Horner, R.L., 2008. Neuromodulation of hypoglossal motoneurons during sleep. Respir Physiol Neurobiol 164, 179-196.

162

Horner, R.L., 2011. Respiratory physiology: central neural control of respiratory neurons and motoneurons during sleep, In: Kryger, M.H., Roth, T., Dement, W.C. (Eds.), Principles and practice of sleep medicine., 5th ed. Elselvier Saunders, St. Louis, Missouri, pp. 237-249.

Horner, R.L., Hughes, S.W., Malhotra, A., 2014. State-dependent and reflex drives to the upper airway: basic physiology with clinical implications. J Appl Physiol 116, 325-336.

Horner, R.L., Innes, J.A., Holden, H.B., Guz, A., 1991a. Afferent pathway(s) for pharyngeal dilator reflex to negative pressure in man: a study using upper airway anaesthesia. J Physiol 436, 31-44.

Horner, R.L., Innes, J.A., Morrell, M.J., Shea, S.A., Guz, A., 1994a. The effect of sleep on reflex genioglossus muscle activation by stimuli of negative airway pressure in humans. J Physiol 476, 141- 151.

Horner, R.L., Innes, J.A., Murphy, K., Guz, A., 1991b. Evidence for reflex upper airway dilator muscle activation by sudden negative airway pressure in man. J Physiol 436, 15-29.

Horner, R.L., Kozar, L.F., Kimoff, R.J., Phillipson, E.A., 1994b. Effects of sleep on the tonic drive to respiratory muscle and the threshold for rhythm generation in the dog. J Physiol 474, 525-537.

Horner, R.L., Shea, S.A., McIvor, J., Guz, A., 1989. Pharyngeal size and shape during wakefulness and sleep in patients with obstructive sleep apnoea. Q J Med 72, 719-735.

Hou, H.M., Hagg, U., Sam, K., Rabie, A.B., Wong, R.W., Lam, B., Ip, M.S., 2006. Dentofacial characteristics of Chinese obstructive sleep apnea patients in relation to obesity and severity. Angle Orthod 76, 962-969.

How, S.C., McConnell, A.K., Taylor, B.J., Romer, L.M., 2007. Acute and chronic responses of the upper airway to inspiratory loading in healthy awake humans: an MRI study. Respir Physiol Neurobiol 157, 270-280.

Hrycyshyn, A.W., Basmajian, J.V., 1972. Electromyography of the oral stage of swallowing in man. Am J Anat 133, 333-340.

Hsu, P.P., Han, H.N., Chan, Y.H., Tay, H.N., Brett, R.H., Lu, P.K., Blair, R.L., 2004. Quantitative computer-assisted digital-imaging upper airway analysis for obstructive sleep apnoea. Clin Otolaryngol Allied Sci 29, 522-529.

Huang, D., Swanson, E.A., Lin, C.P., Schuman, J.S., Stinson, W.G., Chang, W., Hee, M.R., Flotte, T., Gregory, K., Puliafito, C.A., et al., 1991. Optical coherence tomography. Science 254, 1178-1181.

Huang, M.H., Lee, S.T., Rajendran, K., 1997. Structure of the musculus uvulae: functional and surgical implications of an anatomic study. Cleft Palate Craniofac J 34, 466-474.

Hudgel, D.W., 1986. Variable site of airway narrowing among obstructive sleep apnea patients. J Appl Physiol 61, 1403-1409.

Hudgel, D.W., 1992. The role of upper airway anatomy and physiology in obstructive sleep apnea. Clin Chest Med 13, 383-398.

Hudgel, D.W., 1993. Properties of the upper airway during sleep. Ear Nose Throat J 72, 42-45.

163

Hudgel, D.W., Harasick, T., 1990. Fluctuation in timing of upper airway and chest wall inspiratory muscle activity in obstructive sleep apnea. J Appl Physiol 69, 443-450.

Hudgel, D.W., Mulholland, M., Hendricks, C., 1987. Neuromuscular and mechanical responses to inspiratory resistive loading during sleep. J Appl Physiol 63, 603-608.

Hudson, A.L., Butler, J.E., Gandevia, S.C., De Troyer, A., 2010. Interplay between the inspiratory and postural functions of the human parasternal intercostal muscles. J Neurophysiol 103, 1622-1629.

Hudson, A.L., Butler, J.E., Gandevia, S.C., De Troyer, A., 2011. Role of the diaphragm in trunk rotation in humans. J Neurophysiol 106, 1622-1628.

Humbert, I.A., Reeder, S.B., Porcaro, E.J., Kays, S.A., Brittain, J.H., Robbins, J., 2008. Simultaneous estimation of tongue volume and fat fraction using IDEAL-FSE. J Magn Reson Imaging 28, 504-508.

Huxley, A.F., 1957. Muscle structure and theories of contraction. Prog Biophys Biophys Chem 7, 255- 318.

Huxley, A.F., Simmons, R.M., 1971. Proposed mechanism of force generation in striated muscle. Nature 233, 533-538.

Hwang, J.C., Bartlett, D., Jr., St John, W.M., 1983a. Characterization of respiratory-modulated activities of hypoglossal motoneurons. J Appl Physiol Respir 55, 793-798.

Hwang, J.C., St John, W.M., Bartlett, D., Jr., 1983b. Respiratory-related hypoglossal nerve activity: influence of anesthetics. J Appl Physiol 55, 785-792.

Iber, C., Berssenbrugge, A., Skatrud, J.B., Dempsey, J.A., 1982. Ventilatory adaptations to resistive loading during wakefulness and non-REM sleep. J Appl Physiol 52, 607-614.

Ikeda, K., Ogura, M., Oshima, T., Suzuki, H., Higano, S., Takahashi, S., Kurosawa, H., Hida, W., Matsuoka, H., Takasaka, T., 2001. Quantitative assessment of the pharyngeal airway by dynamic magnetic resonance imaging in obstructive sleep apnea syndrome. Ann Otol Rhinol Laryngol 110, 183-189.

Ingman, T., Nieminen, T., Hurmerinta, K., 2004. Cephalometric comparison of pharyngeal changes in subjects with upper airway resistance syndrome or obstructive sleep apnoea in upright and supine positions. Eur J Orthod 26, 321-326.

Insana, M.F., R.F., W., Garra, B.S., Brown, D.G., Shawker, T.H., 1986. Analysis of ultrasound image texture via generalized Rician statistics. Optical Engineering 25, 743-748.

Iskander, A., Sanders, I., 2003. Morphological comparison between neonatal and adult human tongues. Ann Otol Rhinol Laryngol 112, 768-776.

Isono, S., 2012. Obesity and obstructive sleep apnoea: mechanisms for increased collapsibility of the passive pharyngeal airway. Respirology 17, 32-42.

Isono, S., Remmers, J.E., Tanaka, A., Sho, Y., Sato, J., Nishino, T., 1997. Anatomy of pharynx in patients with obstructive sleep apnea and in normal subjects. J Appl Physiol 82, 1319-1326.

164

Isono, S., Shimada, A., Tanaka, A., Ishikawa, T., Nishino, T., Konno, A., 2003. Effects of uvulopalatopharyngoplasty on collapsibility of the retropalatal airway in patients with obstructive sleep apnea. Laryngoscope 113, 362-367.

Isono, S., Shimada, A., Tanaka, A., Tagaito, Y., Utsugi, M., Konno, A., Nishino, T., 1999. Efficacy of endoscopic static pressure/area assessment of the passive pharynx in predicting uvulopalatopharyngoplasty outcomes. Laryngoscope 109, 769-774.

Isono, S., Tanaka, A., Sho, Y., Konno, A., Nishino, T., 1995. Advancement of the mandible improves velopharyngeal airway patency. J Appl Physiol 79, 2132-2138.

Ito, M., Kawakami, Y., Ichinose, Y., Fukashiro, S., Fukunaga, T., 1998. Nonisometric behavior of fascicles during isometric contractions of a human muscle. J Appl Physiol 85, 1230-1235.

Jackson, A.C., Butler, J.P., Millet, E.J., Hoppin, F.G., Jr., Dawson, S.V., 1977. Airway geometry by analysis of acoustic pulse response measurements. J Appl Physiol 43, 523-536.

Jacobs, B.L., Azmitia, E.C., 1992. Structure and function of the brain serotonin system. Physiol Rev 72, 165-229.

Jaghagen, E.L., Berggren, D., Isberg, A., 2000. Swallowing dysfunction related to snoring: a videoradiographic study. Acta Otolaryngol 120, 438-443.

Janczewski, W.A., Feldman, J.L., 2006. Distinct rhythm generators for inspiration and expiration in the juvenile rat. J Physiol 570, 407-420.

Jarvinen, T.A., Jozsa, L., Kannus, P., Jarvinen, T.L., Jarvinen, M., 2002. Organization and distribution of intramuscular connective tissue in normal and immobilized skeletal muscles. An immunohistochemical, polarization and scanning electron microscopic study. J Muscle Res Cell Motil 23, 245-254.

Jelev, A., Sood, S., Liu, H., Nolan, P., Horner, R.L., 2001. Microdialysis perfusion of 5-HT into hypoglossal motor nucleus differentially modulates genioglossus activity across natural sleep-wake states in rats. J Physiol 532, 467-481.

Johal, A., Battagel, J.M., Kotecha, B.T., 2005. Sleep nasendoscopy: a diagnostic tool for predicting treatment success with mandibular advancement splints in obstructive sleep apnoea. Eur J Orthod 27, 607-614.

Johns, M.W., 1991. A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep 14, 540-545.

Johns, M.W., 2002. Sleep propensity varies with behaviour and the situation in which it is measured: the concept of somnificity. J Sleep Res 11, 61-67.

Joosten, S.A., Hamza, K., Sands, S., Turton, A., Berger, P., Hamilton, G., 2012. Phenotypes of patients with mild to moderate obstructive sleep apnoea as confirmed by cluster analysis. Respirology 17, 99- 107.

Jordan, A.S., Catcheside, P.G., O'Donoghue, F.J., Saunders, N.A., McEvoy, R.D., 2002. Genioglossus muscle activity at rest and in response to brief hypoxia in healthy men and women. J Appl Physiol 92, 410-417.

165

Jordan, A.S., Wellman, A., Edwards, J.K., Schory, K., Dover, L., MacDonald, M., Patel, S.R., Fogel, R.B., Malhotra, A., White, D.P., 2005. Respiratory control stability and upper airway collapsibility in men and women with obstructive sleep apnea. J Appl Physiol 99, 2020-2027.

Jordan, A.S., Wellman, A., Heinzer, R.C., Lo, Y.L., Schory, K., Dover, L., Gautam, S., Malhotra, A., White, D.P., 2007. Mechanisms used to restore ventilation after partial upper airway collapse during sleep in humans. Thorax 62, 861-867.

Jouvet, M., 1967. Neurophysiology of the states of sleep. Physiol Rev 47, 117-177.

Kapur, V.K., Auckley, D.H., Chowdhuri, S., Kuhlmann, D.C., Mehra, R., Ramar, K., Harrod, C.G., 2017. Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med 13, 479-504.

Kasai, T., Bradley, T.D., Friedman, O., Logan, A.G., 2014. Effect of intensified diuretic therapy on overnight rostral fluid shift and obstructive sleep apnoea in patients with uncontrolled hypertension. J Hypertens 32, 673-680.

Kasapoglu, F., Dokuzlar, U., 2007. An unknown anatomical variation of omohyoid muscle. Clin Anat 20, 964-965.

Katsantonis, G.P., Walsh, J.K., 1986. Somnofluoroscopy: its role in the selection of candidates for uvulopalatopharyngoplasty. Otolaryngol Head Neck Surg 94, 56-60.

Kawakami, Y., Ichinose, Y., Fukunaga, T., 1998. Architectural and functional features of human triceps surae muscles during contraction. J Appl Physiol 85, 398-404.

Kay, R.F., Cartmill, M., Balow, M., 1998. The hypoglossal canal and the origin of human vocal behavior. Proc Natl Acad Sci U S A 95, 5417-5419.

Kelsey, C.A., Minifie, F.D., Hixon, T.J., 1969. Applications of ultrasound in speech research. J Speech Hear Res 12, 564-575.

Kerwin, W.S., Prince, J.L., 2000. A k-space analysis of MR tagging. J Magn Reson 142, 313-322.

Kier, W.M., Smith, K.K., 1985. Tongues, tentacles and trunks: the biomechanics of movement in muscular-hydrostats. Zool J Linnean Soc 83, 307 - 324.

Kim, A.M., Keenan, B.T., Jackson, N., Chan, E.L., Staley, B., Poptani, H., Torigian, D.A., Pack, A.I., Schwab, R.J., 2014. Tongue fat and its relationship to obstructive sleep apnea. Sleep 37, 1639-1648.

Kim, J., In, K., Kim, J., You, S., Kang, K., Shim, J., Lee, S., Lee, J., Lee, S., Park, C., Shin, C., 2004. Prevalence of sleep-disordered breathing in middle-aged Korean men and women. Am J Respir Crit Care Med 170, 1108-1113.

Kimoff, R.J., 2007. Upper airway myopathy is important in the pathophysiology of obstructive sleep apnea. J Clin Sleep Med 3, 567-569.

Kimoff, R.J., Sforza, E., Champagne, V., Ofiara, L., Gendron, D., 2001. Upper airway sensation in snoring and obstructive sleep apnea. Am J Respir Crit Care Med 164, 250-255.

166

Kinirons, S.A., Shall, M.S., McClung, J.R., Goldberg, S.J., 2003. Effect of artificial rearing on the contractile properties and myosin heavy chain isoforms of developing rat tongue musculature. J Neurophysiol 90, 120-127.

Kirkness, J.P., Schwartz, A.R., Schneider, H., Punjabi, N.M., Maly, J.J., Laffan, A.M., McGinley, B.M., Magnuson, T., Schweitzer, M., Smith, P.L., Patil, S.P., 2008. Contribution of male sex, age, and obesity to mechanical instability of the upper airway during sleep. J Appl Physiol 104, 1618-1624.

Kjaer, M., 2004. Role of extracellular matrix in adaptation of tendon and skeletal muscle to mechanical loading. Physiol Rev 84, 649-698.

Klawe, J.J., Tafil-Klawe, M., 2003. Age-related response of the genioglossus muscle EMG-activity to hypoxia in humans. J Physiol Pharmacol 54 Suppl 1, 14-19.

Kokawa, T., Saigusa, H., Aino, I., Matsuoka, C., Nakamura, T., Tanuma, K., Yamashita, K., Niimi, S., 2006. Physiological studies of retrusive movements of the human tongue. J Voice 20, 414-422.

Koopman, L.P., Slorach, C., Hui, W., Manlhiot, C., McCrindle, B.W., Friedberg, M.K., Jaeggi, E.T., Mertens, L., 2010. Comparison between different speckle tracking and color tissue Doppler techniques to measure global and regional myocardial deformation in children. J Am Soc Echocardiogr 23, 919-928.

Kremkau, F.W., 2016a. Instruments, Sonography: principles and instruments., 9th ed. Elsevier, St. Louis, Missouri, USA., pp. 73-132.

Kremkau, F.W., 2016b. Ultrasound, Sonography: principles and instruments, 9th ed. Elsevier, St. Louis, Missouri, pp. 13-41.

Kremkau, F.W., Taylor, K.J., 1986. Artifacts in ultrasound imaging. J Ultrasound Med 5, 227-237.

Kubin, L., Davies, R.O., Pack, A.I., 1998. Control of upper airway motoneurons during REM sleep. News Physiol Sci 13, 91-97.

Kubin, L., Kimura, H., Tojima, H., Davies, R.O., Pack, A.I., 1993. Suppression of hypoglossal motoneurons during the carbachol-induced atonia of REM sleep is not caused by fast synaptic inhibition. Brain Res 611, 300-312.

Kubin, L., Reignier, C., Tojima, H., Taguchi, O., Pack, A.I., Davies, R.O., 1994. Changes in serotonin level in the hypoglossal nucleus region during carbachol-induced atonia. Brain Res 645, 291-302.

Kuna, S.T., 2000. Respiratory-related activation and mechanical effects of the pharyngeal constrictor muscles. Respir Physiol 119, 155-161.

Kuna, S.T., Bedi, D.G., Ryckman, C., 1988. Effect of nasal airway positive pressure on upper airway size and configuration. Am J Respir Crit Care Med 138, 969-975.

Kuna, S.T., Remmers, J.E., 1999. Premotor input to hypoglossal motoneurons from Kolliker-Fuse neurons in decerebrate cats. Respir Physiol 117, 85-95.

Kuna, S.T., Smickley, J.S., Vanoye, C.R., 1997. Respiratory-related pharyngeal constrictor muscle activity in normal human adults. Am J Respir Crit Care Med 155, 1991-1999.

Kundra, P., Mishra, S.K., Ramesh, A., 2011. Ultrasound of the airway. Indian J Anaesth 55, 456-462.

167

Kuo, G.P., Torok, C.M., Aygun, N., Zinreich, S.J., 2011. Diagnostic imaging of the upper airway. Proc Am Thorac Soc 8, 40-45.

Kwah, L.K., Pinto, R.Z., Diong, J., Herbert, R.D., 2013. Reliability and validity of ultrasound measurements of muscle fascicle length and pennation in humans: a systematic review. J Appl Physiol 114, 761-769.

Kwan, B.C., Butler, J.E., Hudson, A.L., McKenzie, D.K., Bilston, L.E., Gandevia, S.C., 2014. A novel ultrasound technique to measure genioglossus movement in vivo. J Appl Physiol 117, 556-562.

Kwan, B.C., Butler, J.E., Hudson, A.L., McKenzie, D.K., Bilston, L.E., Gandevia, S.C., 2015. Erratum: a novel ultrasound technique to measure genioglossus movement in vivo. J Appl Physiol 118, 1330.

Lahav, Y., Rosenzweig, E., Heyman, Z., Doljansky, J., Green, A., Dagan, Y., 2009. Tongue base ultrasound: a diagnostic tool for predicting obstructive sleep apnea. Ann Otol Rhinol Laryngol 118, 179-184.

Laine, C.M., Bailey, E.F., 2011. Common synaptic input to the human hypoglossal motor nucleus. J Neurophysiol 105, 380-387.

Lakie, M., Robson, L.G., 1988. Thixotropic changes in human muscle stiffness and the effects of fatigue. Q J Exp Physiol 73, 487-500.

Lakie, M., Walsh, E.G., Wright, G.W., 1984. Resonance at the wrist demonstrated by the use of a torque motor: an instrumental analysis of muscle tone in man. J Physiol 353, 265-285.

Lam, S., Standish, B., Baldwin, C., McWilliams, A., leRiche, J., Gazdar, A., Vitkin, A.I., Yang, V., Ikeda, N., MacAulay, C., 2008. In vivo optical coherence tomography imaging of preinvasive bronchial lesions. Clin Cancer Res 14, 2006-2011.

Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for categorical data. Biometrics 33, 159-174.

Larsson, H., Carlsson-Nordlander, B., Lindblad, L.E., Norbeck, O., Svanborg, E., 1992. Temperature thresholds in the oropharynx of patients with obstructive sleep apnea syndrome. Am J Respir Crit Care Med 146, 1246-1249.

Launois, S., Whitelaw, W.A., 1990. The upper airway and respiratory disorders during sleep. Rev Mal Respir 7, 391-402.

Launois, S.H., Feroah, T.R., Campbell, W.N., Issa, F.G., Morrison, D., Whitelaw, W.A., Isono, S., Remmers, J.E., 1993. Site of pharyngeal narrowing predicts outcome of surgery for obstructive sleep apnea. Am J Respir Crit Care Med 147, 182-189.

168

Lavie, P., Pillar, G., Malhotra, A., 2002. Sleep disorders: diagnosis, management and treatment. A handbook for clinicians. Taylor & Francis, Florence, Kentucky.

Leiter, J.C., 1996. Upper airway shape: Is it important in the pathogenesis of obstructive sleep apnea? Am J Respir Crit Care Med 153, 894-898.

Li, H.Y., Chen, N.H., Wang, C.R., Shu, Y.H., Wang, P.C., 2003a. Use of 3-dimensional computed tomography scan to evaluate upper airway patency for patients undergoing sleep-disordered breathing surgery. Otolaryngol Head Neck Surg 129, 336-342.

Li, K.K., Kushida, C., Powell, N.B., Riley, R.W., Guilleminault, C., 2000. Obstructive sleep apnea syndrome: a comparison between Far-East Asian and white men. Laryngoscope 110, 1689-1693.

Li, M., Kambhamettu, C., Stone, M., 2005. Tongue motion averaging from contour sequences. Clin Linguist Phon 19, 515-528.

Li, Y.M., Shen, L., Peever, J.H., Duffin, J., 2003b. Connections between respiratory neurones in the neonatal rat transverse medullary slice studied with cross-correlation. J Physiol 549, 327-332.

Liistro, G., Stanescu, D., Dooms, G., Rodenstein, D., Veriter, C., 1988. Head position modifies upper airway resistance in men. J Appl Physiol 64, 1285-1288.

Lindman, R., Stal, P.S., 2002. Abnormal palatopharyngeal muscle morphology in sleep-disordered breathing. J Neurol Sci 195, 11-23.

Liu, Y., Lowe, A.A., Zeng, X., Fu, M., Fleetham, J.A., 2000. Cephalometric comparisons between Chinese and Caucasian patients with obstructive sleep apnea. Am J Orthod Dentofacial Orthop 117, 479-485.

Lo, Y.L., Jordan, A.S., Malhotra, A., Wellman, A., Heinzer, R.C., Schory, K., Dover, L., Fogel, R.B., White, D.P., 2006. Genioglossal muscle response to CO2 stimulation during NREM sleep. Sleep 29, 470-477.

Loewen, A.H., Ostrowski, M., Laprairie, J., Maturino, F., Hanly, P.J., Younes, M., 2011. Response of genioglossus muscle to increasing chemical drive in sleeping obstructive apnea patients. Sleep 34, 1061-1073.

Loram, I.D., Maganaris, C.N., Lakie, M., 2006. Use of ultrasound to make noninvasive in vivo measurement of continuous changes in human muscle contractile length. J Appl Physiol 100, 1311- 1323.

Lowe, A.A., 1980. The neural regulation of tongue movements. Prog Neurobiol 15, 295-344.

Lowe, A.A., Gionhaku, N., Takeuchi, K., Fleetham, J.A., 1986. Three-dimensional CT reconstructions of tongue and airway in adult subjects with obstructive sleep apnea. Am J Orthod Dentofacial Orthop 90, 364-374.

Ludemann, P., Dziewas, R., Soros, P., Happe, S., Frese, A., 2001. Axonal polyneuropathy in obstructive sleep apnoea. J Neurol Neurosurg Psychiatry 70, 685-687.

Luu, B.L., Muceli, S., Saboisky, J.P., Farina, D., Heroux, M.E., Bilston, L.E., Gandevia, S.C., Butler, J.E., 2017. Motor unit territories in human genioglossus estimated with multi-channel intramuscular electrodes. J Appl Physiol.

169

Lyberg, T., Krogstad, O., Djupesland, G., 1989a. Cephalometric analysis in patients with obstructive sleep apnoea syndrome: II. Soft tissue morphology. J Laryngol Otol 103, 293-297.

Lyberg, T., Krogstad, O., Djupesland, G., 1989b. Cephalometric analysis in patients with obstructive sleep apnoea syndrome. I. Skeletal morphology. J Laryngol Otol 103, 287-292.

Macefield, G., Gandevia, S.C., 1991. The cortical drive to human respiratory muscles in the awake state assessed by premotor cerebral potentials. J Physiol 439, 545-558.

Maganaris, C.N., 2001. Force-length characteristics of in vivo human skeletal muscle. Acta Physiol Scand 172, 279-285.

Maganaris, C.N., Baltzopoulos, V., Sargeant, A.J., 1998. In vivo measurements of the triceps surae complex architecture in man: implications for muscle function. J Physiol 512 ( Pt 2), 603-614.

Magnusson, S.P., Aagaard, P., Dyhre-Poulsen, P., Kjaer, M., 2001. Load-displacement properties of the human triceps surae aponeurosis in vivo. J Physiol 531, 277-288.

Magnusson, S.P., Hansen, P., Aagaard, P., Brond, J., Dyhre-Poulsen, P., Bojsen-Moller, J., Kjaer, M., 2003. Differential strain patterns of the human gastrocnemius aponeurosis and free tendon, in vivo. Acta Physiol Scand 177, 185-195.

Malhotra, A., Fogel, R.B., Edwards, J.K., Shea, S.A., White, D.P., 2000a. Local mechanisms drive genioglossus activation in obstructive sleep apnea. Am J Respir Crit Care Med 161, 1746-1749.

Malhotra, A., Huang, Y., Fogel, R., Lazic, S., Pillar, G., Jakab, M., Kikinis, R., White, D.P., 2006. Aging influences on pharyngeal anatomy and physiology: the predisposition to pharyngeal collapse. Am J Med 119, 72 e79-14.

Malhotra, A., Pillar, G., Fogel, R.B., Beauregard, J., Edwards, J.K., Slamowitz, D.I., Shea, S.A., White, D.P., 2000b. Genioglossal but not palatal muscle activity relates closely to pharyngeal pressure. Am J Respir Crit Care Med 162, 1058-1062.

Malhotra, A., Pillar, G., Fogel, R.B., Edwards, J.K., Ayas, N., Akahoshi, T., Hess, D., White, D.P., 2002. Pharyngeal pressure and flow effects on genioglossus activation in normal subjects. Am J Respir Crit Care Med 165, 71-77.

Malhotra, A., White, D.P., 2002. Obstructive sleep apnoea. Lancet 360, 237-245.

Manaker, S., Tischler, L.J., 1993. Origin of serotoninergic afferents to the hypoglossal nucleus in the rat. J Comp Neurol 334, 466-476.

Marcus, C.L., Lutz, J., Hamer, A., Smith, P.L., Schwartz, A., 1999. Developmental changes in response to subatmospheric pressure loading of the upper airway. J Appl Physiol 87, 626-633.

Martelli, D., Stanic, D., Dutschmann, M., 2013. The emerging role of the parabrachial complex in the generation of wakefulness drive and its implication for respiratory control. Respir Physiol Neurobiol 188, 318-323.

Martin, S.E., Mathur, R., Marshall, I., Douglas, N.J., 1997. The effect of age, sex, obesity and posture on upper airway size. Eur Respir J 10, 2087-2090.

Martin-Harris, B., 2006. PART 1 Oral cavity, pharynx and esophagus. GI Motility online.

170

Marwick, T.H., 2010. Consistency of myocardial deformation imaging between vendors. Eur J Echocardiogr 11, 414-416.

Mason, M., Cates, C.J., Smith, I., 2015. Effects of opioid, hypnotic and sedating medications on sleep- disordered breathing in adults with obstructive sleep apnoea. Cochrane Database Syst Rev, CD011090.

Mateika, J.H., Millrood, D.L., Kim, J., Rodriguez, H.P., Samara, G.J., 1999. Response of human tongue protrudor and retractors to hypoxia and hypercapnia. Am J Respir Crit Care Med 160, 1976-1982.

Mathew, O.P., Abu-Osba, Y.K., Thach, B.T., 1982. Genioglossus muscle responses to upper airway pressure changes: afferent pathways. J Appl Physiol 52, 445-450.

Mathur, R., Douglas, N.J., 1995. Family studies in patients with the sleep apnea-hypopnea syndrome. Ann Intern Med 122, 174-178.

Mayer, P., Dematteis, M., Pepin, J.L., Wuyam, B., Veale, D., Vila, A., Levy, P., 1999. Peripheral neuropathy in sleep apnea. A tissue marker of the severity of nocturnal desaturation. Am J Respir Crit Care Med 159, 213-219.

Mayer, P., Pepin, J.L., Bettega, G., Veale, D., Ferretti, G., Deschaux, C., Levy, P., 1996. Relationship between body mass index, age and upper airway measurements in snorers and sleep apnoea patients. Eur Respir J 9, 1801-1809.

McClung, J.R., Goldberg, S.J., 1999. Organization of motoneurons in the dorsal hypoglossal nucleus that innervate the retrusor muscles of the tongue in the rat. Anat Rec 254, 222-230.

McClung, J.R., Goldberg, S.J., 2000. Functional anatomy of the hypoglossal innervated muscles of the rat tongue: a model for elongation and protrusion of the mammalian tongue. Anat Rec 260, 378-386.

McClung, J.R., Goldberg, S.J., 2002. Organization of the hypoglossal motoneurons that innervate the horizontal and oblique components of the genioglossus muscle in the rat. Brain Res 950, 321-324.

McDicken, W.N., 1991. Diagnostic Ultrasonics. Principles and Use of Instruments, 3 ed. Churchill Livingstone, New York.

McDicken, W.N., Anderson, T., 2011. Basic physics of medical ultrasound, Clinical ultrasound, 3rd ed. Churchill Livingstone, Spain, pp. 3-15.

McGinty, D., Szymusiak, R., 2011. Neural control of sleep in mammals, In: Siegel, J.M. (Ed.), Principles and practice of sleep medicine, 5th ed. Elservier Saunders, St. Louis, Missouri, pp. 76-91.

McGraw, K.O., Wong, S.P., 1996. Forming inferences about some intraclass correlation coefficients. Psychological Methods 1, 30-46.

McLaughlin, R.A., Noble, P.B., Sampson, D.D., 2014. Optical coherence tomography in respiratory science and medicine: from airways to alveoli. Physiology (Bethesda) 29, 369-380.

Mellen, N.M., Janczewski, W.A., Bocchiaro, C.M., Feldman, J.L., 2003. Opioid-induced quantal slowing reveals dual networks for respiratory rhythm generation. Neuron 37, 821-826.

171

Merida-Velasco, J.R., Rodriguez-Vazquez, J.F., de la Cuadra Blanco, C., Sanchez-Montesinos, I., Merida-Velasco, J.A., 2006. Origin of the styloglossus muscle in the human fetus. J Anat 208, 649- 653.

Mezzanotte, W.S., Tangel, D.J., White, D.P., 1992. Waking genioglossal electromyogram in sleep apnea patients versus normal controls (a neuromuscular compensatory mechanism). J Clin Invest 89, 1571-1579.

Middleton, W.D., Kurtz, A.B., Hertzberg, B.S., 2004. Practical physics, Ultrasound, the Requisites, 2nd ed. Mosby, Boston, pp. 3-27.

Miki, H., Hida, W., Chonan, T., Kikuchi, Y., Takishima, T., 1989. Effects of submental electrical stimulation during sleep on upper airway patency in patients with obstructive sleep apnea. Am J Respir Crit Care Med 140, 1285-1289.

Miller, J.L., Kang, S.M., 2007. Preliminary ultrasound observation of lingual movement patterns during nutritive versus non-nutritive sucking in a premature infant. Dysphagia 22, 150-160.

Miller, J.L., Watkin, K.L., Chen, M.F., 2002. Muscle, adipose, and connective tissue variations in intrinsic musculature of the adult human tongue. J Speech Lang Hear Res 45, 51-65.

Minifie, F.D., 1973. Speech acoustics, In: Minifie, F.D., Hixon, T.J., Williams, F.E. (Eds.), Normal aspects of speech, hearing and language. Prentice-Hall, Englewood Cliffs, NJ, pp. 235-284.

Montagnat, J., Delingette, H., 2005. 4D deformable models with temporal constraints: application to 4D cardiac image segmentation. Med Image Anal 9, 87-100.

Morrell, M.J., Arabi, Y., Zahn, B., Badr, M.S., 1998. Progressive retropalatal narrowing preceding obstructive apnea. Am J Respir Crit Care Med 158, 1974-1981.

Mortimore, I.L., Bennett, S.P., Douglas, N.J., 2000. Tongue protrusion strength and fatiguability: relationship to apnoea/hypopnoea index and age. J Sleep Res 9, 389-393.

Mortimore, I.L., Douglas, N.J., 1996. Palatopharyngeus has respiratory activity and responds to negative pressure in sleep apnoeics. Eur Respir J 9, 773-778.

Mortimore, I.L., Douglas, N.J., 1997. Palatal muscle EMG response to negative pressure in awake sleep apneic and control subjects. Am J Respir Crit Care Med 156, 867-873.

Moruzzi, G., 1972. The sleep-waking cycle. Ergeb Physiol 64, 1-165.

Mu, L., Sanders, I., 2000. Neuromuscular specializations of the pharyngeal dilator muscles: II. Compartmentalization of the canine genioglossus muscle. Anat Rec 260, 308-325.

Mu, L., Sanders, I., 2010. Human tongue neuroanatomy: Nerve supply and motor endplates. Clin Anat 23, 777-791.

Murphy, K., Mier, A., Adams, L., Guz, A., 1990. Putative cerebral cortical involvement in the ventilatory response to inhaled CO2 in conscious man. J Physiol 420, 1-18.

Muthupillai, R., Lomas, D.J., Rossman, P.J., Greenleaf, J.F., Manduca, A., Ehman, R.L., 1995. Magnetic resonance elastography by direct visualization of propagating acoustic strain waves. Science 269, 1854-1857.

172

Mwenge, G.B., Rombaux, P., Dury, M., Lengele, B., Rodenstein, D., 2013. Targeted hypoglossal neurostimulation for obstructive sleep apnoea: a 1-year pilot study. Eur Respir J 41, 360-367.

Mwenge, G.B., Rombaux, P., Lengele, B., Rodenstein, D., 2015. Hypoglossal nerve stimulation for obstructive sleep apnea. Prog Neurol Surg 29, 94-105.

Nakayama, K., Niwa, M., Sasaki, S.I., Ichikawa, T., Hirai, N., 1998. Morphology of single primary spindle afferents of the intercostal muscles in the cat. J Comp Neurol 398, 459-472.

Napadow, V.J., Chen, Q., Wedeen, V.J., Gilbert, R.J., 1999a. Biomechanical basis for lingual muscular deformation during swallowing. Am J Physiol 277, G695-701.

Napadow, V.J., Chen, Q., Wedeen, V.J., Gilbert, R.J., 1999b. Intramural mechanics of the human tongue in association with physiological deformations. J Biomech 32, 1-12.

Narici, M.V., Binzoni, T., Hiltbrand, E., Fasel, J., Terrier, F., Cerretelli, P., 1996. In vivo human gastrocnemius architecture with changing joint angle at rest and during graded isometric contraction. J Physiol 496 ( Pt 1), 287-297.

Nashi, N., Kang, S., Barkdull, G.C., Lucas, J., Davidson, T.M., 2007. Lingual fat at autopsy. Laryngoscope 117, 1467-1473.

Nathan, P.W., 1963. The descending respiratory pathway in man. J Neurol Neurosurg Psychiatry 26, 487-499.

Nauta, W.J., 1946. Hypothalamic regulation of sleep in rats; an experimental study. J Neurophysiol 9, 285-316.

Nelson, S., Hans, M., 1997. Contribution of craniofacial risk factors in increasing apneic activity among obese and nonobese habitual snorers. Chest 111, 154-162.

Netzer, N.C., Stoohs, R.A., Netzer, C.M., Clark, K., Strohl, K.P., 1999. Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med 131, 485-491.

Nguyen, A.T., Jobin, V., Payne, R., Beauregard, J., Naor, N., Kimoff, R.J., 2005. Laryngeal and velopharyngeal sensory impairment in obstructive sleep apnea. Sleep 28, 585-593.

Nicholas, C.L., Bei, B., Worsnop, C., Malhotra, A., Jordan, A.S., Saboisky, J.P., Chan, J.K., Duckworth, E., White, D.P., Trinder, J., 2010. Motor unit recruitment in human genioglossus muscle in response to hypercapnia. Sleep 33, 1529-1538.

Niitsu, M., Kumada, M., Campeau, N.G., Niimi, S., Riederer, S.J., Itai, Y., 1994. Tongue displacement: visualization with rapid tagged magnetization-prepared MR imaging. Radiology 191, 578-580.

Nishikawa, K.C., Kier, W.M., Smith, K.K., 1999. Morphology and mechanics of tongue movement in the African pig-nosed frog Hemisus marmoratum: a muscular hydrostatic model. J Exp Biol 202, 771- 780.

Nishio, J., Matsuya, T., Machida, J., Miyazaki, T., 1976. The motor nerve supply of the velopharyngeal muscles. Cleft Palate J 13, 20-30.

Nunnelly, J.C., 1978. Psychometric Theory, 2 ed. McGraw Hill, New York.

173

O'Dwyer, N.J., Quinn, P.T., Guitar, B.E., Andrews, G., Neilson, P.D., 1981. Procedures for verification of electrode placement in EMG studies of orofacial and mandibular muscles. J Speech Hear Res 24, 273-288.

O'Reilly, P.M., FitzGerald, M.J., 1990. Fibre composition of the hypoglossal nerve in the rat. J Anat 172, 227-243.

Ogata, S., Mine, K., Tamatsu, Y., Shimada, K., 2002. Morphological study of the human chondroglossus muscle in Japanese. Ann Anat 184, 493-499.

Okada, T., Fukatsu, H., Ishigaki, T., Yasuma, F., Kayukawa, Y., 1996. Ultra-low-field magnetic resonance imaging in upper airways obstruction in sleep apnea syndrome. Psychiatry Clin Neurosci 50, 285-289.

Oliven, A., O'Hearn, D.J., Boudewyns, A., Odeh, M., De Backer, W., van de Heyning, P., Smith, P.L., Eisele, D.W., Allan, L., Schneider, H., Testerman, R., Schwartz, A.R., 2003. Upper airway response to electrical stimulation of the genioglossus in obstructive sleep apnea. J Appl Physiol 95, 2023-2029.

Oliven, A., Odeh, M., Geitini, L., Oliven, R., Steinfeld, U., Schwartz, A.R., Tov, N., 2007. Effect of coactivation of tongue protrusor and retractor muscles on pharyngeal lumen and airflow in sleep apnea patients. J Appl Physiol 103, 1662-1668.

Onal, E., Lopata, M., O'Connor, T.D., 1981. Diaphragmatic and genioglossal electromyogram responses to isocapnic hypoxia in humans. Am J Respir Crit Care Med 124, 215-217.

Onimaru, H., Homma, I., 2006. Point:Counterpoint: The parafacial respiratory group (pFRG)/pre- Botzinger complex (preBotC) is the primary site of respiratory rhythm generation in the mammal. Point: the PFRG is the primary site of respiratory rhythm generation in the mammal. J Appl Physiol 100, 2094-2095.

Ono, T., Ishiwata, Y., Inaba, N., Kuroda, T., Nakamura, Y., 1994. Hypoglossal premotor neurons with rhythmical inspiratory-related activity in the cat: localization and projection to the phrenic nucleus. Exp Brain Res 98, 1-12.

Orban, P., Rauchs, G., Balteau, E., Degueldre, C., Luxen, A., Maquet, P., Peigneux, P., 2006. Sleep after spatial learning promotes covert reorganization of brain activity. Proc Natl Acad Sci U S A 103, 7124-7129.

Orem, J., Kubin, L., 2000. Respiratory physiology: central neural control., In: Kryger, M.H., Roth, T., Dement, W.C. (Eds.), Principles and Practice of Sleep Medicine, 3rd ed. WB Saunders, Philadelphia, pp. 205-220.

Orem, J., Lovering, A.T., Dunin-Barkowski, W., Vidruk, E.H., 2000. Endogenous excitatory drive to the respiratory system in rapid eye movement sleep in cats. J Physiol 527 Pt 2, 365-376.

Osman, N.F., Kerwin, W.S., McVeigh, E.R., Prince, J.L., 1999. Cardiac motion tracking using CINE harmonic phase (HARP) magnetic resonance imaging. Magn Reson Med 42, 1048-1060.

Otonari-Yamamoto, M., Nakajima, K., Tsuji, Y., Otonari, T., Curtin, H.D., Okano, T., Sano, T., 2010. Imaging of the mylohyoid muscle: separation of submandibular and sublingual spaces. Am J Roentgenol 194, W431-438.

174

Owens, R.L., Edwards, B.A., Sands, S.A., Butler, J.P., Eckert, D.J., White, D.P., Malhotra, A., Wellman, A., 2014. The classical Starling resistor model often does not predict inspiratory airflow patterns in the human upper airway. J Appl Physiol 116, 1105-1112.

Pae, E.K., Hyatt, J.P., Wu, J., Chien, P., 2007. Short-term electrical stimulation alters tongue muscle fibre type composition. Arch Oral Biol 52, 544-551.

Pae, E.K., Lowe, A.A., Fleetham, J.A., 1997. A role of pharyngeal length in obstructive sleep apnea patients. Am J Orthod Dentofacial Orthop 111, 12-17.

Pappas, G.P., Asakawa, D.S., Delp, S.L., Zajac, F.E., Drace, J.E., 2002. Nonuniform shortening in the biceps brachii during elbow flexion. J Appl Physiol 92, 2381-2389.

Parisi, R.A., Neubauer, J.A., Frank, M.M., Edelman, N.H., Santiago, T.V., 1987. Correlation between genioglossal and diaphragmatic responses to hypercapnia during sleep. Am J Respir Crit Care Med 135, 378-382.

Partinen, M., Guilleminault, C., Quera-Salva, M.A., Jamieson, A., 1988. Obstructive sleep apnea and cephalometric roentgenograms. The role of anatomic upper airway abnormalities in the definition of abnormal breathing during sleep. Chest 93, 1199-1205.

Pataka, A., Daskalopoulou, E., Kalamaras, G., Fekete Passa, K., Argyropoulou, P., 2014. Evaluation of five different questionnaires for assessing sleep apnea syndrome in a sleep clinic. Sleep Med 15, 776- 781.

Patrick, G.B., Strohl, K.P., Rubin, S.B., Altose, M.D., 1982. Upper airway and diaphragm muscle responses to chemical stimulation and loading. J Appl Physiol 53, 1133-1137.

Peigneux, P., Laureys, S., Delbeuck, X., Maquet, P., 2001. Sleeping brain, learning brain. The role of sleep for memory systems. Neuroreport 12, A111-124.

Peng, C.L., Jost-Brinkmann, P.G., Yoshida, N., Miethke, R.R., Lin, C.T., 2003. Differential diagnosis between infantile and mature swallowing with ultrasonography. Eur J Orthod 25, 451-456.

Peng, C.L., Miethke, R.R., Pong, S.J., Lin, C.T., 2007. Investigation of tongue movements during swallowing with M-mode ultrasonography. J Orofac Orthop 68, 17-25.

Pengo, M.F., Xiao, S., Ratneswaran, C., Reed, K., Shah, N., Chen, T., Douiri, A., Hart, N., Luo, Y., Rafferty, G.F., Rossi, G.P., Williams, A., Polkey, M.I., Moxham, J., Steier, J., 2016. Randomised sham- controlled trial of transcutaneous electrical stimulation in obstructive sleep apnoea. Thorax 71, 923- 931.

Peppard, P.E., Young, T., Barnet, J.H., Palta, M., Hagen, E.W., Hla, K.M., 2013. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol 177, 1006-1014.

Peppard, P.E., Young, T., Palta, M., Dempsey, J., Skatrud, J., 2000a. Longitudinal study of moderate weight change and sleep-disordered breathing. JAMA 284, 3015-3021.

Peppard, P.E., Young, T., Palta, M., Skatrud, J., 2000b. Prospective study of the association between sleep-disordered breathing and hypertension. N Engl J Med 342, 1378-1384.

Petrof, B.J., Hendricks, J.C., Pack, A.I., 1996. Does upper airway muscle injury trigger a vicious cycle in obstructive sleep apnea? A hypothesis. Sleep 19, 465-471.

175

Pette, D., Staron, R.S., 1997. Mammalian skeletal muscle fiber type transitions. Int Rev Cytol 170, 143-223.

Pillar, G., Fogel, R.B., Malhotra, A., Beauregard, J., Edwards, J.K., Shea, S.A., White, D.P., 2001. Genioglossal inspiratory activation: central respiratory vs mechanoreceptive influences. Respir Physiol 127, 23-38.

Pillar, G., Malhotra, A., Fogel, R., Beauregard, J., Schnall, R., White, D.P., 2000a. Airway mechanics and ventilation in response to resistive loading during sleep: influence of gender. Am J Respir Crit Care Med 162, 1627-1632.

Pillar, G., Malhotra, A., Fogel, R.B., Beauregard, J., Slamowitz, D.I., Shea, S.A., White, D.P., 2000b. Upper airway muscle responsiveness to rising PCO(2) during NREM sleep. J Appl Physiol 89, 1275- 1282.

Pittayapat, P., Limchaichana-Bolstad, N., Willems, G., Jacobs, R., 2014. Three-dimensional cephalometric analysis in orthodontics: a systematic review. Orthod Craniofac Res 17, 69-91.

Pittman, L.J., Bailey, E.F., 2009. Genioglossus and intrinsic electromyographic activities in impeded and unimpeded protrusion tasks. J Neurophysiol 101, 276-282.

Podoleanu, A.G., 2005. Optical coherence tomography. Br J Radiol 78, 976-988.

Popovic, R.M., White, D.P., 1995. Influence of gender on waking genioglossal electromyogram and upper airway resistance. Am J Respir Crit Care Med 152, 725-731.

Popratiloff, A.S., Streppel, M., Gruart, A., Guntinas-Lichius, O., Angelov, D.N., Stennert, E., Delgado- Garcia, J.M., Neiss, W.F., 2001. Hypoglossal and reticular interneurons involved in oro-facial coordination in the rat. J Comp Neurol 433, 364-379.

Pracharktam, N., Hans, M.G., Strohl, K.P., Redline, S., 1994. Upright and supine cephalometric evaluation of obstructive sleep apnea syndrome and snoring subjects. Angle Orthod 64, 63-73.

Prasad, A., Yu, E., Wong, D.T., Karkhanis, R., Gullane, P., Chan, V.W., 2011. Comparison of sonography and computed tomography as imaging tools for assessment of airway structures. J Ultrasound Med 30, 965-972.

Prince, J.L., McVeigh, E.R., 1992. Motion estimation from tagged MR image sequences. IEEE Trans Med Imaging 11, 238-249.

Proske, U., Morgan, D.L., Gregory, J.E., 1993. Thixotropy in skeletal muscle and in muscle spindles: a review. Prog Neurobiol 41, 705-721.

Punjabi, N.M., 2008. The epidemiology of adult obstructive sleep apnea. Proc Am Thorac Soc 5, 136- 143.

Quintana-Gallego, E., Carmona-Bernal, C., Capote, F., Sanchez-Armengol, A., Botebol-Benhamou, G., Polo-Padillo, J., Castillo-Gomez, J., 2004. Gender differences in obstructive sleep apnea syndrome: a clinical study of 1166 patients. Respir Med 98, 984-989.

Rack, P.M., Westbury, D.R., 1974. The short range stiffness of active mammalian muscle and its effect on mechanical properties. J Physiol 240, 331-350.

176

Ralph, M.R., Foster, R.G., Davis, F.C., Menaker, M., 1990. Transplanted suprachiasmatic nucleus determines circadian period. Science 247, 975-978.

Ramchandren, S., Gruis, K.L., Chervin, R.D., Lisabeth, L.D., Concannon, M., Wolfe, J., Albers, J.W., Brown, D.L., 2010. Hypoglossal nerve conduction findings in obstructive sleep apnea. Muscle Nerve 42, 257-261.

Rasch, B., Buchel, C., Gais, S., Born, J., 2007. Odor cues during slow-wave sleep prompt declarative memory consolidation. Science 315, 1426-1429.

Rauchs, G., Orban, P., Schmidt, C., Albouy, G., Balteau, E., Degueldre, C., Schnackers, C., Sterpenich, V., Tinguely, G., Luxen, A., Maquet, P., Peigneux, P., 2008. Sleep modulates the neural substrates of both spatial and contextual memory consolidation. PLoS One 3, e2949.

Redline, S., Kirchner, H.L., Quan, S.F., Gottlieb, D.J., Kapur, V., Newman, A., 2004. The effects of age, sex, ethnicity, and sleep-disordered breathing on sleep architecture. Arch Intern Med 164, 406-418.

Redline, S., Tishler, P.V., Hans, M.G., Tosteson, T.D., Strohl, K.P., Spry, K., 1997. Racial differences in sleep-disordered breathing in African-Americans and Caucasians. Am J Respir Crit Care Med 155, 186-192.

Redline, S., Tishler, P.V., Tosteson, T.D., Williamson, J., Kump, K., Browner, I., Ferrette, V., Krejci, P., 1995. The familial aggregation of obstructive sleep apnea. Am J Respir Crit Care Med 151, 682-687.

Redline, S., Tosteson, T., Tishler, P.V., Carskadon, M.A., Millman, R.P., 1992. Studies in the genetics of obstructive sleep apnea. Familial aggregation of symptoms associated with sleep-related breathing disturbances. Am J Respir Crit Care Med 145, 440-444.

Redolfi, S., Arnulf, I., Pottier, M., Lajou, J., Koskas, I., Bradley, T.D., Similowski, T., 2011. Attenuation of obstructive sleep apnea by compression stockings in subjects with venous insufficiency. Am J Respir Crit Care Med 184, 1062-1066.

Redolfi, S., Yumino, D., Ruttanaumpawan, P., Yau, B., Su, M.C., Lam, J., Bradley, T.D., 2009. Relationship between overnight rostral fluid shift and obstructive sleep apnea in nonobese men. Am J Respir Crit Care Med 179, 241-246.

Remmers, J.E., deGroot, W.J., Sauerland, E.K., Anch, A.M., 1978. Pathogenesis of upper airway occlusion during sleep. J Appl Physiol 44, 931-938.

Rhys, R., 2011. Ultrasound of the neck, In: Allan, P., Baxter, G., Weston, M. (Eds.), Clinical Ultrasound, 3 ed. Churchill Livingstone, China, pp. 890 - 919.

Rikard-Bell, G.C., Bystrzycka, E.K., Nail, B.S., 1985. The identification of brainstem neurones projecting to thoracic respiratory motoneurones in the cat as demonstrated by retrograde transport of HRP. Brain Res Bull 14, 25-37.

Riley, R., Guilleminault, C., Herran, J., Powell, N., 1983. Cephalometric analyses and flow-volume loops in obstructive sleep apnea patients. Sleep 6, 303-311.

Riley, R., Guilleminault, C., Powell, N., Simmons, F.B., 1985. Palatopharyngoplasty failure, cephalometric roentgenograms, and obstructive sleep apnea. Otolaryngol Head Neck Surg 93, 240- 244.

177

Ritter, C.T., Trudo, F.J., Goldberg, A.N., Welch, K.C., Maislin, G., Schwab, R.J., 1999. Quantitative evaluation of the upper airway during nasopharyngoscopy with the Muller maneuver. Laryngoscope 109, 954-963.

Rivlin, J., Hoffstein, V., Kalbfleisch, J., McNicholas, W., Zamel, N., Bryan, A.C., 1984. Upper airway morphology in patients with idiopathic obstructive sleep apnea. Am J Respir Crit Care Med 129, 355- 360.

Rodenstein, D.O., Dooms, G., Thomas, Y., Liistro, G., Stanescu, D.C., Culee, C., Aubert-Tulkens, G., 1990. Pharyngeal shape and dimensions in healthy subjects, snorers, and patients with obstructive sleep apnoea. Thorax 45, 722-727.

Rodenstein, D.O., Stanescu, D.C., 1984. Soft palate and oronasal breathing in humans. J Appl Physiol 57, 651-657.

Rojewski, T.E., Schuller, D.E., Clark, R.W., Schmidt, H.S., Potts, R.E., 1982. Synchronous video recording of the pharyngeal airway and polysomnograph in patients with obstructive sleep apnea. Laryngoscope 92, 246-250.

Rowe, R.W., 1981. Morphology of perimysial and endomysial connective tissue in skeletal muscle. Tissue Cell 13, 681-690.

Rowley, J.A., Permutt, S., Willey, S., Smith, P.L., Schwartz, A.R., 1996. Effect of tracheal and tongue displacement on upper airway airflow dynamics. J Appl Physiol 80, 2171-2178.

Rubinstein, I., Hoffstein, V., Bradley, T.D., 1989. Lung volume-related changes in the pharyngeal area of obese females with and without obstructive sleep apnoea. Eur Respir J 2, 344-351.

Rutherford, O.M., Jones, D.A., 1992. Measurement of fibre pennation using ultrasound in the human quadriceps in vivo. Eur J Appl Physiol Occup Physiol 65, 433-437.

Ryan, C.F., Love, L.L., 1996. Mechanical properties of the velopharynx in obese patients with obstructive sleep apnea. Am J Respir Crit Care Med 154, 806-812.

Ryan, C.F., Love, L.L., 2000. Unpredictable results of laser assisted uvulopalatoplasty in the treatment of obstructive sleep apnoea. Thorax 55, 399-404.

Ryan, C.F., Lowe, A.A., Li, D., Fleetham, J.A., 1991a. Magnetic resonance imaging of the upper airway in obstructive sleep apnea before and after chronic nasal continuous positive airway pressure therapy. Am J Respir Crit Care Med 144, 939-944.

Ryan, C.F., Lowe, A.A., Li, D., Fleetham, J.A., 1991b. Three-dimensional upper airway computed tomography in obstructive sleep apnea. A prospective study in patients treated by uvulopalatopharyngoplasty. Am J Respir Crit Care Med 144, 428-432.

Ryan, W.J., Hawkins, C.F., 1976. Ultrasonic measurement of lateral pharyngeal wall movement at the velopharyngeal port. Cleft Palate J 13, 156-164.

Saboisky, J., 2008. Neural drive to human respiratory muscles, Prince of Wales Medical Research Institute. University of New South Wales, Sydney.

178

Saboisky, J.P., Butler, J.E., Fogel, R.B., Taylor, J.L., Trinder, J.A., White, D.P., Gandevia, S.C., 2006. Tonic and phasic respiratory drives to human genioglossus motoneurons during breathing. J Neurophysiol 95, 2213-2221.

Saboisky, J.P., Butler, J.E., Luu, B.L., Gandevia, S.C., 2015. Neurogenic changes in the upper airway of obstructive sleep apnoea. Curr Neurol Neurosci Rep 15, 12.

Saboisky, J.P., Butler, J.E., McKenzie, D.K., Gorman, R.B., Trinder, J.A., White, D.P., Gandevia, S.C., 2007a. Neural drive to human genioglossus in obstructive sleep apnoea. J Physiol 585, 135-146.

Saboisky, J.P., Butler, J.E., Walsh, L.D., Gandevia, S.C., 2007b. New display of the timing and firing frequency of single motor units. J Neurosci Methods 162, 287-292.

Saboisky, J.P., Gorman, R.B., De Troyer, A., Gandevia, S.C., Butler, J.E., 2007c. Differential activation among five human inspiratory motoneuron pools during tidal breathing. J Appl Physiol 102, 772-780.

Saboisky, J.P., Jordan, A.S., Eckert, D.J., White, D.P., Trinder, J.A., Nicholas, C.L., Gautam, S., Malhotra, A., 2010. Recruitment and rate-coding strategies of the human genioglossus muscle. J Appl Physiol 109, 1939-1949.

Saboisky, J.P., Luu, B., Butler, J.E., Gandevia, S.C., 2013. Optimal length for tongue protrusion force in humans., European Respiratory Society Barcelona, p. 976s.

Saboisky, J.P., Stashuk, D.W., Hamilton-Wright, A., Carusona, A.L., Campana, L.M., Trinder, J., Eckert, D.J., Jordan, A.S., McSharry, D.G., White, D.P., Nandedkar, S., David, W.S., Malhotra, A., 2012. Neurogenic changes in the upper airway of patients with obstructive sleep apnea. Am J Respir Crit Care Med 185, 322-329.

Safar, P., Escarraga, L.A., Chang, F., 1959. Upper airway obstruction in the unconscious patient. J Appl Physiol 14, 760-764.

Saigusa, H., Niimi, S., Yamashita, K., Gotoh, T., Kumada, M., 2001. Morphological and histochemical studies of the genioglossus muscle. Ann Otol Rhinol Laryngol 110, 779-784.

Saigusa, H., Saigusa, M., Aino, I., Iwasaki, C., Li, L., Niimi, S., 2006a. M-mode color Doppler ultrasonic imaging of vertical tongue movement during articulatory movement. J Voice 20, 38-45.

Saigusa, H., Tanuma, K., Yamashita, K., Saigusa, M., Niimi, S., 2006b. Nerve fiber analysis for the lingual nerve of the human adult subjects. Surg Radiol Anat 28, 59-65.

Saigusa, H., Yamashita, K., Tanuma, K., Saigusa, M., Niimi, S., 2004. Morphological studies for retrusive movement of the human adult tongue. Clin Anat 17, 93-98.

Saito, H., Itoh, I., 2003. Three-dimensional architecture of the intrinsic tongue muscles, particularly the longitudinal muscle, by the chemical-maceration method. Anat Sci Int 78, 168-176.

Saito, H., Itoh, I., 2007. The three-dimensional architecture of the human styloglossus especially its posterior muscle bundles. Ann Anat 189, 261-267.

Salame, K., Masharawi, Y., Rochkind, S., Arensburg, B., 2006. Surgical anatomy of the cervical segment of the hypoglossal nerve. Clin Anat 19, 37-43.

179

Sanders, I., Mu, L., 2013. A three-dimensional atlas of human tongue muscles. Anat Rec 296, 1102- 1114.

Sanders, I., Mu, L., Amirali, A., Su, H., Sobotka, S., 2013. The human tongue slows down to speak: muscle fibers of the human tongue. Anat Rec 296, 1615-1627.

Sanders, M.H., Kern, N., 1990. Obstructive sleep apnea treated by independently adjusted inspiratory and expiratory positive airway pressures via nasal mask. Physiologic and clinical implications. Chest 98, 317-324.

Sanders, M.H., Moore, S.E., 1983. Inspiratory and expiratory partitioning of airway resistance during sleep in patients with sleep apnea. Am J Respir Crit Care Med 127, 554-558.

Sauerland, E.K., Harper, R.M., 1976. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Exp Neurol 51, 160-170.

Sauerland, E.K., Mitchell, S.P., 1975. Electromyographic activity of intrinsic and extrinsic muscles of the human tongue. Tex Rep Biol Med 33, 444-455.

Sawczuk, A., Mosier, K.M., 2001. Neural control of tongue movement with respect to respiration and swallowing. Crit Rev Oral Biol Med 12, 18-37.

Schnall, R.P., Pillar, G., Kelsen, S.G., Oliven, A., 1995. Dilatory effects of upper airway muscle contraction induced by electrical stimulation in awake humans. J Appl Physiol 78, 1950-1956.

Schwab, R.J., 1996. Properties of tissues surrounding the upper airway. Sleep 19, S170-174.

Schwab, R.J., 1998. Upper airway imaging. Clin Chest Med 19, 33-54.

Schwab, R.J., 2001. Imaging for the snoring and sleep apnea patient. Dent Clin North Am 45, 759- 796.

Schwab, R.J., Gefter, W.B., Hoffman, E.A., Gupta, K.B., Pack, A.I., 1993. Dynamic upper airway imaging during awake respiration in normal subjects and patients with sleep disordered breathing. Am J Respir Crit Care Med 148, 1385-1400.

Schwab, R.J., Goldberg, A.N., 1998. Upper airway assessment: radiographic and other imaging techniques. Otolaryngol Clin North Am 31, 931-968.

Schwab, R.J., Gupta, K.B., Gefter, W.B., Metzger, L.J., Hoffman, E.A., Pack, A.I., 1995. Upper airway and soft tissue anatomy in normal subjects and patients with sleep-disordered breathing. Significance of the lateral pharyngeal walls. Am J Respir Crit Care Med 152, 1673-1689.

Schwab, R.J., Pack, A.I., Gupta, K.B., Metzger, L.J., Oh, E., Getsy, J.E., Hoffman, E.A., Gefter, W.B., 1996. Upper airway and soft tissue structural changes induced by CPAP in normal subjects. Am J Respir Crit Care Med 154, 1106-1116.

Schwab, R.J., Pasirstein, M., Pierson, R., Mackley, A., Hachadoorian, R., Arens, R., Maislin, G., Pack, A.I., 2003. Identification of upper airway anatomic risk factors for obstructive sleep apnea with volumetric magnetic resonance imaging. Am J Respir Crit Care Med 168, 522-530.

180

Schwartz, A.R., Barnes, M., Hillman, D., Malhotra, A., Kezirian, E., Smith, P.L., Hoegh, T., Parrish, D., Eastwood, P.R., 2012. Acute upper airway responses to hypoglossal nerve stimulation during sleep in obstructive sleep apnea. Am J Respir Crit Care Med 185, 420-426.

Schwartz, A.R., Eisele, D.W., Hari, A., Testerman, R., Erickson, D., Smith, P.L., 1996. Electrical stimulation of the lingual musculature in obstructive sleep apnea. J Appl Physiol 81, 643-652.

Schwartz, A.R., Gold, A.R., Schubert, N., Stryzak, A., Wise, R.A., Permutt, S., Smith, P.L., 1991. Effect of weight loss on upper airway collapsibility in obstructive sleep apnea. Am J Respir Crit Care Med 144, 494-498.

Schwartz, A.R., Smith, P.L., Oliven, A., 2014. Electrical stimulation of the hypoglossal nerve: a potential therapy. J Appl Physiol 116, 337-344.

Schwartz, A.R., Smith, P.L., Wise, R.A., Gold, A.R., Permutt, S., 1988. Induction of upper airway occlusion in sleeping individuals with subatmospheric nasal pressure. J Appl Physiol 64, 535-542.

Senaratna, C.V., Perret, J.L., Lodge, C.J., Lowe, A.J., Campbell, B.E., Matheson, M.C., Hamilton, G.S., Dharmage, S.C., 2017. Prevalence of obstructive sleep apnea in the general population: A systematic review. Sleep Med Rev 34, 70-81.

Series, F., 2002. Upper airway muscles awake and asleep. Sleep Med Rev 6, 229-242.

Series, F., Cormier, Y., Desmeules, M., 1990. Influence of passive changes of lung volume on upper airways. J Appl Physiol 68, 2159-2164.

Series, F., Cote, C., Simoneau, J.A., Gelinas, Y., St Pierre, S., Leclerc, J., Ferland, R., Marc, I., 1995. Physiologic, metabolic, and muscle fiber type characteristics of musculus uvulae in sleep apnea hypopnea syndrome and in snorers. J Clin Invest 95, 20-25.

Series, F.J., Simoneau, S.A., St Pierre, S., Marc, I., 1996. Characteristics of the genioglossus and musculus uvulae in sleep apnea hypopnea syndrome and in snorers. Am J Respir Crit Care Med 153, 1870-1874.

Sforza, E., Bacon, W., Weiss, T., Thibault, A., Petiau, C., Krieger, J., 2000. Upper airway collapsibility and cephalometric variables in patients with obstructive sleep apnea. Am J Respir Crit Care Med 161, 347-352.

Sha, B.F., England, S.J., Parisi, R.A., Strobel, R.J., 2000. Force production of the genioglossus as a function of muscle length in normal humans. J Appl Physiol 88, 1678-1684.

Sharshar, T., Hopkinson, N.S., Jonville, S., Prigent, H., Carlier, R., Dayer, M.J., Swallow, E.B., Lofaso, F., Moxham, J., Polkey, M.I., 2004. Demonstration of a second rapidly conducting cortico-diaphragmatic pathway in humans. J Physiol 560, 897-908.

Shawker, T.H., Sonies, B., Stone, M., Baum, B.J., 1983. Real-time ultrasound visualization of tongue movement during swallowing. J Clin Ultrasound 11, 485-490.

Shawker, T.H., Sonies, B.C., 1984. Tongue movement during speech: a real-time ultrasound evaluation. J Clin Ultrasound 12, 125-133.

Shawker, T.H., Sonies, B.C., Stone, M., 1984. Soft tissue anatomy of the tongue and floor of the mouth: an ultrasound demonstration. Brain Lang 21, 335-350.

181

Shea, S.A., Edwards, J.K., White, D.P., 1999. Effect of wake-sleep transitions and rapid eye movement sleep on pharyngeal muscle response to negative pressure in humans. J Physiol 520 Pt 3, 897-908.

Shelton, K.E., Gay, S.B., Hollowell, D.E., Woodson, H., Suratt, P.M., 1993a. Mandible enclosure of upper airway and weight in obstructive sleep apnea. Am J Respir Crit Care Med 148, 195-200.

Shelton, K.E., Woodson, H., Gay, S., Suratt, P.M., 1993b. Pharyngeal fat in obstructive sleep apnea. Am J Respir Crit Care Med 148, 462-466.

Shepard, J.W., Jr., Garrison, M., Vas, W., 1990a. Upper airway distensibility and collapsibility in patients with obstructive sleep apnea. Chest 98, 84-91.

Shepard, J.W., Jr., Gefter, W.B., Guilleminault, C., Hoffman, E.A., Hoffstein, V., Hudgel, D.W., Suratt, P.M., White, D.P., 1991. Evaluation of the upper airway in patients with obstructive sleep apnea. Sleep 14, 361-371.

Shepard, J.W., Jr., Stanson, A.W., Sheedy, P.F., Westbrook, P.R., 1990b. Fast-CT evaluation of the upper airway during wakefulness in patients with obstructive sleep apnea. Prog Clin Biol Res 345, 273-279; discussion 280-272.

Shepard, J.W., Jr., Thawley, S.E., 1989. Evaluation of the upper airway by computerized tomography in patients undergoing uvulopalatopharyngoplasty for obstructive sleep apnea. Am J Respir Crit Care Med 140, 711-716.

Shepherd, K.L., Jensen, C.M., Maddison, K.J., Hillman, D.R., Eastwood, P.R., 2006. Relationship between upper airway and inspiratory pump muscle force in obstructive sleep apnea. Chest 130, 1757-1764.

Sher, A.E., Thorpy, M.J., Shprintzen, R.J., Spielman, A.J., Burack, B., McGregor, P.A., 1985. Predictive value of Muller maneuver in selection of patients for uvulopalatopharyngoplasty. Laryngoscope 95, 1483-1487.

Shiina, T., Nightingale, K.R., Palmeri, M.L., Hall, T.J., Bamber, J.C., Barr, R.G., Castera, L., Choi, B.I., Chou, Y.H., Cosgrove, D., Dietrich, C.F., Ding, H., Amy, D., Farrokh, A., Ferraioli, G., Filice, C., Friedrich- Rust, M., Nakashima, K., Schafer, F., Sporea, I., Suzuki, S., Wilson, S., Kudo, M., 2015. WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 1: basic principles and terminology. Ultrasound Med Biol 41, 1126-1147.

Shingaki, M., Nikkuni, Y., Katsura, K., Ikeda, N., Maruyama, S., Takagi, R., Hayashi, T., 2016. Clinical significance of intraoral strain elastography for diagnosing early stage tongue carcinoma: a preliminary study. Oral Radiology.

Shintani, T., Anker, A.R., Billig, I., Card, J.P., Yates, B.J., 2003. Transneuronal tracing of neural pathways influencing both diaphragm and genioglossal muscle activity in the ferret. J Appl Physiol 95, 1453-1459.

Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: uses in assessing rater reliability. Psychol Bull 86, 420-428.

Shu, C.C., Lee, P., Lin, J.W., Huang, C.T., Chang, Y.C., Yu, C.J., Wang, H.C., 2013. The use of sub-mental ultrasonography for identifying patients with severe obstructive sleep apnea. PLoS One 8, e62848.

182

Siegel, H., Sonies, B.C., Graham, B., McCutchen, C., Hunter, K., Vega-Bermudez, F., Sato, S., 2000. Obstructive sleep apnea: A study by simultaneous polysomnography and ultrasonic imaging. Neurology 54, 1872.

Siegel, J.M., 2005. Clues to the functions of mammalian sleep. Nature 437, 1264-1271.

Sigrist, R.M.S., Liau, J., Kaffas, A.E., Chammas, M.C., Willmann, J.K., 2017. Ultrasound Elastography: Review of Techniques and Clinical Applications. Theranostics 7, 1303-1329.

Silverstein, K., Costello, B.J., Giannakpoulos, H., Hendler, B., 2000. Genioglossus muscle attachments: an anatomic analysis and the implications for genioglossus advancement. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 90, 686-688.

Singh, M., Chin, K.J., Chan, V.W., Wong, D.T., Prasad, G.A., Yu, E., 2010. Use of sonography for airway assessment: an observational study. J Ultrasound Med 29, 79-85.

Skolnick, M.L., Zagzebski, J.A., Watkin, K.L., 1975. Two dimensional ultrasonic demonstration of lateral pharyngeal wall movement in real time--a preliminary report. Cleft Palate J 12, 299-303.

Slaughter, K., Li, H., Sokoloff, A.J., 2005. Neuromuscular organization of the superior longitudinalis muscle in the human tongue. 1. Motor endplate morphology and muscle fiber architecture. Cells Tissues Organs 181, 51-64.

Smith, C., 2001. Sleep states and memory processes in humans: procedural versus declarative memory systems. Sleep Med Rev 5, 491-506.

Smith, J.C., Ellenberger, H.H., Ballanyi, K., Richter, D.W., Feldman, J.L., 1991. Pre-Botzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254, 726-729.

Smith, J.C., Goldberg, S.J., Shall, M.S., 2005. Phenotype and contractile properties of mammalian tongue muscles innervated by the hypoglossal nerve. Respir Physiol Neurobiol 147, 253-262.

Smith, K., Kier, W., 1989. Trunks, tongues, and tentacles: moving with skeletons of muscle. American Scientist 77, 28-35.

Smith, P.L., Wise, R.A., Gold, A.R., Schwartz, A.R., Permutt, S., 1988. Upper airway pressure-flow relationships in obstructive sleep apnea. J Appl Physiol 64, 789-795.

Smith, S.W., Sandrik, J.M., Wagner, R.F., van Ramm, O.T., 1982. Measurements and Analysis of Speckle in Ultrasound B-Scans, In: Alais, P., Metherell, A.F. (Eds.), Acoustical Imaging. Springer US, Boston, MA, pp. 195-211.

Sokoloff, A., Burkholder, T., 2012. Tongue structure and function, In: McLoon, L., Andrade, F. (Eds.), Craniofacial muscles. Springer, New York.

Sokoloff, A.J., 2000. Localization and contractile properties of intrinsic longitudinal motor units of the rat tongue. J Neurophysiol 84, 827-835.

Sokoloff, A.J., Daugherty, M., Li, H., 2010. Myosin heavy-chain composition of the human hyoglossus muscle. Dysphagia 25, 81-93.

Sokoloff, A.J., Deacon, T.W., 1992. Musculotopic organization of the hypoglossal nucleus in the cynomolgus monkey, Macaca fascicularis. J Comp Neurol 324, 81-93.

183

Sokoloff, A.J., Yang, B., Li, H., Burkholder, T.J., 2007. Immunohistochemical characterization of slow and fast myosin heavy chain composition of muscle fibres in the styloglossus muscle of the human and macaque (Macaca rhesus). Arch Oral Biol 52, 533-543.

Sonies, B.C., Shawker, T.H., Hall, T.E., Gerber, L.H., Leighton, S.B., 1981. Ultrasonic visualization of tongue motion during speech. J Acoust Soc Am 70, 683-686.

Sonntag, C.F., 1925. The Comparative Anatomy of the Tongues of the Mammalia.—XII. Summary, Classification and Phylogeny. Proceedings of the Zoological Society of London 95, 701-762.

Spann, R.W., Hyatt, R.E., 1971. Factors affecting upper airway resistance in conscious man. J Appl Physiol 31, 708-712.

Spiegel, K., Leproult, R., Colecchia, E.F., L'Hermite-Baleriaux, M., Nie, Z., Copinschi, G., Van Cauter, E., 2000. Adaptation of the 24-h growth hormone profile to a state of sleep debt. Am J Physiol Regul Integr Comp Physiol 279, R874-883.

Spiegel, K., Leproult, R., L'Hermite-Baleriaux, M., Copinschi, G., Penev, P.D., Van Cauter, E., 2004. Leptin levels are dependent on sleep duration: relationships with sympathovagal balance, carbohydrate regulation, cortisol, and thyrotropin. J Clin Endocrinol Metab 89, 5762-5771.

Spiegel, K., Leproult, R., Van Cauter, E., 1999. Impact of sleep debt on metabolic and endocrine function. Lancet 354, 1435-1439.

Spiro, J., Rendell, J.K., Gay, T., 1994. Activation and coordination patterns of the suprahyoid muscles during swallowing. Laryngoscope 104, 1376-1382.

St-John, W.M., Paton, J.F., Leiter, J.C., 2004. Uncoupling of rhythmic hypoglossal from phrenic activity in the rat. Exp Physiol 89, 727-737.

Stal, P., Marklund, S., Thornell, L.E., De Paul, R., Eriksson, P.O., 2003. Fibre composition of human intrinsic tongue muscles. Cells Tissues Organs 173, 147-161.

Stanchina, M.L., Malhotra, A., Fogel, R.B., Trinder, J., Edwards, J.K., Schory, K., White, D.P., 2003. The influence of lung volume on pharyngeal mechanics, collapsibility, and genioglossus muscle activation during sleep. Sleep 26, 851-856.

Stavness, I., Hannam, A.G., Lloyd, J.E., Fels, S., 2008. Towards predicting biomechanical consequences of jaw reconstruction. Conf Proc IEEE Eng Med Biol Soc 2008, 4567-4570.

Steier, J., Seymour, J., Rafferty, G.F., Jolley, C.J., Solomon, E., Luo, Y., Man, W.D., Polkey, M.I., Moxham, J., 2011. Continuous transcutaneous submental electrical stimulation in obstructive sleep apnea: a feasibility study. Chest 140, 998-1007.

Stein, M.G., Gamsu, G., de Geer, G., Golden, J.A., Crumley, R.L., Webb, W.R., 1987. Cine CT in obstructive sleep apnea. Am J Roentgenol 148, 1069-1074.

Stickgold, R., Walker, M.P., 2007. Sleep-dependent memory consolidation and reconsolidation. Sleep Med 8, 331-343.

Stockx, E.M., Camilleri, P., Skuza, E.M., Churchward, T., Howes, J.M., Ho, M., McDonald, T., Freezer, N., Hamilton, G., Wilkinson, M.H., Berger, P.J., 2010. New acoustic method for detecting upper airway obstruction in patients with sleep apnoea. Respirology 15, 326-335.

184

Stone, M., 2005. A guide to analysing tongue motion from ultrasound images. Clin Linguist Phon 19, 455-501.

Stone, M., Davis, E.P., 1995. A head and transducer support system for making ultrasound images of tongue/jaw movement. J Acoust Soc Am 98, 3107-3112.

Stone, M., Davis, E.P., Douglas, A.S., NessAiver, M., Gullapalli, R., Levine, W.S., Lundberg, A., 2001. Modeling the motion of the internal tongue from tagged cine-MRI images. J Acoust Soc Am 109, 2974-2982.

Stone, M., Shawker, T.H., Talbot, T.L., Rich, A.H., 1988. Cross-sectional tongue shape during the production of vowels. J Acoust Soc Am 83, 1586-1596.

Stone, M., Woo, J., Lee, J., Poole, T., Seagraves, A., Chung, M., Kim, E., Murano, E.Z., Prince, J.L., Blemker, S.S., 2016. Structure and variability in human tongue muscle anatomy. Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 1-9.

Strobel, R.J., Rosen, R.C., 1996. Obesity and weight loss in obstructive sleep apnea: a critical review. Sleep 19, 104-115.

Strohl, K.P., Hensley, M.J., Hallett, M., Saunders, N.A., Ingram, R.H., Jr., 1980. Activation of upper airway muscles before onset of inspiration in normal humans. J Appl Physiol 49, 638-642.

Strollo, P.J., Jr., Soose, R.J., Maurer, J.T., de Vries, N., Cornelius, J., Froymovich, O., Hanson, R.D., Padhya, T.A., Steward, D.L., Gillespie, M.B., Woodson, B.T., Van de Heyning, P.H., Goetting, M.G., Vanderveken, O.M., Feldman, N., Knaack, L., Strohl, K.P., Group, S.T., 2014. Upper-airway stimulation for obstructive sleep apnea. N Engl J Med 370, 139-149.

Stuck, B.A., Maurer, J.T., 2008. Airway evaluation in obstructive sleep apnea. Sleep Med Rev 12, 411- 436.

Sunnergren, O., Brostrom, A., Svanborg, E., 2011. Soft palate sensory neuropathy in the pathogenesis of obstructive sleep apnea. Laryngoscope 121, 451-456.

Suratt, P.M., Dee, P., Atkinson, R.L., Armstrong, P., Wilhoit, S.C., 1983. Fluoroscopic and computed tomographic features of the pharyngeal airway in obstructive sleep apnea. Am J Respir Crit Care Med 127, 487-492.

Suratt, P.M., McTier, R.F., Wilhoit, S.C., 1988. Upper airway muscle activation is augmented in patients with obstructive sleep apnea compared with that in normal subjects. Am J Respir Crit Care Med 137, 889-894.

Suratt, P.M., Wilhoit, S.C., Cooper, K., 1984. Induction of airway collapse with subatmospheric pressure in awake patients with sleep apnea. J Appl Physiol 57, 140-146.

Sustic, A., 2007. Role of ultrasound in the airway management of critically ill patients. Crit Care Med 35, S173-177.

Suto, Y., Matsuo, T., Kato, T., Hori, I., Inoue, Y., Ogawa, S., Suzuki, T., Yamada, M., Ohta, Y., 1993. Evaluation of the pharyngeal airway in patients with sleep apnea: value of ultrafast MR imaging. Am J Roentgenol 160, 311-314.

185

Suzue, T., 1984. Respiratory rhythm generation in the in vitro brain stem-spinal cord preparation of the neonatal rat. J Physiol 354, 173-183.

Svanborg, E., 2005. Impact of obstructive apnea syndrome on upper airway respiratory muscles. Respir Physiol Neurobiol 147, 263-272.

Takahashi, S., Ono, T., Ishiwata, Y., Kuroda, T., 2002. Breathing modes, body positions, and suprahyoid muscle activity. J Orthod 29, 307-313; discussion 279.

Takemoto, H., 2001. Morphological analyses of the human tongue musculature for three- dimensional modeling. J Speech Lang Hear Res 44, 95-107.

Tangel, D.J., Mezzanotte, W.S., Sandberg, E.J., White, D.P., 1992. Influences of NREM sleep on the activity of tonic vs. inspiratory phasic muscles in normal men. J Appl Physiol 73, 1058-1066.

Teramoto, S., Sudo, E., Matsuse, T., Ohga, E., Ishii, T., Ouchi, Y., Fukuchi, Y., 1999. Impaired swallowing reflex in patients with obstructive sleep apnea syndrome. Chest 116, 17-21.

Thut, D.C., Schwartz, A.R., Roach, D., Wise, R.A., Permutt, S., Smith, P.L., 1993. Tracheal and neck position influence upper airway airflow dynamics by altering airway length. J Appl Physiol 75, 2084- 2090.

Togeiro, S.M., Chaves, C.M., Jr., Palombini, L., Tufik, S., Hora, F., Nery, L.E., 2010. Evaluation of the upper airway in obstructive sleep apnoea. Indian J Med Res 131, 230-235.

Tomlins, P.H., Wang, R.K., 2005. Theory, developments and applications of optical coherence tomography. J Phys D 38, 2519.

Travers, J.B., Jackson, L.M., 1992. Hypoglossal neural activity during licking and swallowing in the awake rat. J Neurophysiol 67, 1171-1184.

Trinder, J., Jordan, A.S., Nicholas, C.L., 2014. Discharge properties of upper airway motor units during wakefulness and sleep. Prog Brain Res 212, 59-75.

Trinder, J., Kay, A., Kleiman, J., Dunai, J., 1997. Gender differences in airway resistance during sleep. J Appl Physiol 83, 1986-1997.

Trudo, F.J., Gefter, W.B., Welch, K.C., Gupta, K.B., Maislin, G., Schwab, R.J., 1998. State-related changes in upper airway caliber and surrounding soft-tissue structures in normal subjects. Am J Respir Crit Care Med 158, 1259-1270.

Tsuboi, M., Hayashi, A., Ikeda, N., Honda, H., Kato, Y., Ichinose, S., Kato, H., 2005. Optical coherence tomography in the diagnosis of bronchial lesions. Lung Cancer 49, 387-394.

Tsui, B.C., Hui, C.M., 2008. Sublingual airway ultrasound imaging. Can J Anaesth 55, 790-791.

Tsui, B.C., Hui, C.M., 2009. Challenges in sublingual airway ultrasound interpretation. Can J Anaesth 56, 393-394.

Tsuiki, S., Ono, T., Ishiwata, Y., Kuroda, T., 2000. Functional divergence of human genioglossus motor units with respiratory-related activity. Eur Respir J 15, 906-910.

186

Tufik, S., Santos-Silva, R., Taddei, J.A., Bittencourt, L.R., 2010. Obstructive sleep apnea syndrome in the Sao Paulo Epidemiologic Sleep Study. Sleep Med 11, 441-446.

Van de Graaff, W.B., 1988. Thoracic influence on upper airway patency. J Appl Physiol 65, 2124- 2131.

Van de Graaff, W.B., 1991. Thoracic traction on the trachea: mechanisms and magnitude. J Appl Physiol 70, 1328-1336.

Van de Heyning, P.H., Badr, M.S., Baskin, J.Z., Cramer Bornemann, M.A., De Backer, W.A., Dotan, Y., Hohenhorst, W., Knaack, L., Lin, H.S., Maurer, J.T., Netzer, A., Odland, R.M., Oliven, A., Strohl, K.P., Vanderveken, O.M., Verbraecken, J., Woodson, B.T., 2012. Implanted upper airway stimulation device for obstructive sleep apnea. Laryngoscope 122, 1626-1633.

Van Eijden, T.M., Korfage, J.A., Brugman, P., 1997. Architecture of the human jaw-closing and jaw- opening muscles. Anat Rec 248, 464-474. van Lunteren, E., Strohl, K.P., 1986. The muscles of the upper airways. Clin Chest Med 7, 171-188.

Van Lunteren, E., Van de Graaff, W.B., Parker, D.M., Strohl, K.P., Mitra, J., Salamone, J., Cherniack, N.S., 1983. Activity of upper airway muscles during augmented breaths. Respir Physiol 53, 87-98.

Vitti, M., Basmajian, J.V., 1977. Integrated actions of masticatory muscles: simultaneous EMG from eight intramuscular electrodes. Anat Rec 187, 173-189. von Economo, C., 1930. Sleep as a problem of localization. Journal of nervous and mental disease. 71, 249-259.

Vranish, J.R., Bailey, E.F., 2015. A comprehensive assessment of genioglossus electromyographic activity in healthy adults. J Neurophysiol 113, 2692-2699.

Wagner, R.F., Insana, M.F., Brown, D.G., 1986. Unified approach to the detection and classification of speckle texture in diagnostic ultrasound. Opt Eng 25, 738-742.

Walsh, J.H., Leigh, M.S., Paduch, A., Maddison, K.J., Philippe, D.L., Armstrong, J.J., Sampson, D.D., Hillman, D.R., Eastwood, P.R., 2008. Evaluation of pharyngeal shape and size using anatomical optical coherence tomography in individuals with and without obstructive sleep apnoea. J Sleep Res 17, 230-238.

Wang, J., Han, Y., Su, H., Mu, L., 2004. Expression of unique and developmental myosin heavy chain isoforms in adult human digastric muscle. J Histochem Cytochem 52, 851-859.

Wasicko, M.J., Hutt, D.A., Parisi, R.A., Neubauer, J.A., Mezrich, R., Edelman, N.H., 1990. The role of vascular tone in the control of upper airway collapsibility. Am J Respir Crit Care Med 141, 1569-1577.

Wasicko, M.J., Leiter, J.C., Erlichman, J.S., Strobel, R.J., Bartlett, D., Jr., 1991. Nasal and pharyngeal resistance after topical mucosal vasoconstriction in normal humans. Am J Respir Crit Care Med 144, 1048-1052.

Watkin, K.L., Zagzebski, J.A., 1973. Letter: On-line ultrasonic technique for monitoring tongue displacements. J Acoust Soc Am 54, 544-547.

187

Weissengruber, G.E., Forstenpointner, G., Peters, G., Kubber-Heiss, A., Fitch, W.T., 2002. Hyoid apparatus and pharynx in the lion (Panthera leo), jaguar (Panthera onca), tiger (Panthera tigris), cheetah (Acinonyxjubatus) and domestic cat (Felis silvestris f. catus). J Anat 201, 195-209.

Welch, K.C., Foster, G.D., Ritter, C.T., Wadden, T.A., Arens, R., Maislin, G., Schwab, R.J., 2002. A novel volumetric magnetic resonance imaging paradigm to study upper airway anatomy. Sleep 25, 532- 542.

Wellman, A., Genta, P.R., Owens, R.L., Edwards, B.A., Sands, S.A., Loring, S.H., White, D.P., Jackson, A.C., Pedersen, O.F., Butler, J.P., 2014. Test of the Starling resistor model in the human upper airway during sleep. J Appl Physiol 117, 1478-1485.

Wellman, A., Jordan, A.S., Malhotra, A., Fogel, R.B., Katz, E.S., Schory, K., Edwards, J.K., White, D.P., 2004. Ventilatory control and airway anatomy in obstructive sleep apnea. Am J Respir Crit Care Med 170, 1225-1232.

Wellman, A., Malhotra, A., Jordan, A.S., Schory, K., Gautam, S., White, D.P., 2007. Chemical control stability in the elderly. J Physiol 581, 291-298.

Wetter, D.W., Young, T.B., Bidwell, T.R., Badr, M.S., Palta, M., 1994. Smoking as a risk factor for sleep-disordered breathing. Arch Intern Med 154, 2219-2224.

Wheatley, J.R., Amis, T.C., Engel, L.A., 1991. Influence of nasal airflow temperature and pressure on alae nasi electrical activity. J Appl Physiol 71, 2283-2291.

Wheatley, J.R., Mezzanotte, W.S., Tangel, D.J., White, D.P., 1993a. Influence of sleep on genioglossus muscle activation by negative pressure in normal men. Am J Respir Crit Care Med 148, 597-605.

Wheatley, J.R., Tangel, D.J., Mezzanotte, W.S., White, D.P., 1993b. Influence of sleep on response to negative airway pressure of tensor palatini muscle and retropalatal airway. J Appl Physiol 75, 2117- 2124.

White, D.P., 1986. Occlusion pressure and ventilation during sleep in normal humans. J Appl Physiol 61, 1279-1287.

White, D.P., Douglas, N.J., Pickett, C.K., Weil, J.V., Zwillich, C.W., 1982. Hypoxic ventilatory response during sleep in normal premenopausal women. Am J Respir Crit Care Med 126, 530-533.

White, D.P., Lombard, R.M., Cadieux, R.J., Zwillich, C.W., 1985. Pharyngeal resistance in normal humans: influence of gender, age, and obesity. J Appl Physiol 58, 365-371.

White, S.R., 1985. A comparison of the effects of serotonin, substance P and thyrotropin-releasing hormone on excitability of rat spinal motoneurons in vivo. Brain Res 335, 63-70.

Wiegand, D.A., Latz, B., Zwillich, C.W., Wiegand, L., 1990. Upper airway resistance and geniohyoid muscle activity in normal men during wakefulness and sleep. J Appl Physiol 69, 1252-1261.

Wilkinson, V., Malhotra, A., Nicholas, C.L., Worsnop, C., Jordan, A.S., Butler, J.E., Saboisky, J.P., Gandevia, S.C., White, D.P., Trinder, J., 2008. Discharge patterns of human genioglossus motor units during sleep onset. Sleep 31, 525-533.

188

Wilkinson, V., Malhotra, A., Nicholas, C.L., Worsnop, C., Jordan, A.S., Butler, J.E., Saboisky, J.P., Gandevia, S.C., White, D.P., Trinder, J., 2010. Discharge patterns of human genioglossus motor units during arousal from sleep. Sleep 33, 379-387.

Will, M.J., Ester, M.S., Ramirez, S.G., Tiner, B.D., McAnear, J.T., Epstein, L., 1995. Comparison of cephalometric analysis with ethnicity in obstructive sleep apnea syndrome. Sleep 18, 873-875.

Williams, H.L., Holloway, F.A., Griffiths, W.J., 1973. Physiological psychology: sleep. Annu Rev Psychol 24, 279-316.

Williamson, J.P., Armstrong, J.J., McLaughlin, R.A., Noble, P.B., West, A.R., Becker, S., Curatolo, A., Noffsinger, W.J., Mitchell, H.W., Phillips, M.J., Sampson, D.D., Hillman, D.R., Eastwood, P.R., 2010. Measuring airway dimensions during bronchoscopy using anatomical optical coherence tomography. Eur Respir J 35, 34-41.

Woch, G., Davies, R.O., Pack, A.I., Kubin, L., 1996. Behaviour of raphe cells projecting to the dorsomedial medulla during carbachol-induced atonia in the cat. J Physiol 490 ( Pt 3), 745-758.

Wojtczak, J.A., 2012. Submandibular sonography: assessment of hyomental distances and ratio, tongue size, and floor of the mouth musculature using portable sonography. J Ultrasound Med 31, 523-528.

Woo, J., Lee, J., Murano, E.Z., Xing, F., Al-Talib, M., Stone, M., Prince, J.L., 2015. A high-resolution atlas and statistical model of the vocal tract from structural MRI. Comput Methods Biomech Biomed Eng Imaging Vis 3, 47-60.

Woo, J., Murano, E.Z., Stone, M., Prince, J.L., 2012. Reconstruction of high-resolution tongue volumes from MRI. IEEE Trans Biomed Eng 59, 3511-3524.

Woodson, B.T., Garancis, J.C., Toohill, R.J., 1991. Histopathologic changes in snoring and obstructive sleep apnea syndrome. Laryngoscope 101, 1318-1322.

Woodson, B.T., Wooten, M.R., 1994. Comparison of upper-airway evaluations during wakefulness and sleep. Laryngoscope 104, 821-828.

Worsnop, C., Kay, A., Pierce, R., Kim, Y., Trinder, J., 1998. Activity of respiratory pump and upper airway muscles during sleep onset. J Appl Physiol 85, 908-920.

Xing, F., Prince, J.L., Stone, M., Wedeen, V.J., Fakhri, G.E., Woo, J., 2017. A Four-dimensional Motion Field Atlas of the Tongue from Tagged and Cine Magnetic Resonance Imaging. Proc SPIE Int Soc Opt Eng 10133.

Xu, C., Brennick, M.J., Dougherty, L., Wootton, D.M., 2009. Modeling upper airway collapse by a finite element model with regional tissue properties. Med Eng Phys 31, 1343-1348.

Yagel, S., Cohen, S.M., Shapiro, I., Valsky, D.V., 2007. 3D and 4D ultrasound in fetal cardiac scanning: a new look at the fetal heart. Ultrasound Obstet Gynecol 29, 81-95.

Yang, C.S., Stone, M., 2002. Dynamic programming method for remporal registration of three- dimensional tongue surface motion from multiple utterances. Speech Commun 38, 199-207.

Yasui, Y., Kogo, M., Iida, S., Hamaguchi, M., Koizumi, H., Kohara, H., Matsuya, T., 1993. Respiratory activities in relation to external glossal muscles. J Osaka Univ Dent Sch 33, 27-33.

189

Yildirim, N., Fitzpatrick, M.F., Whyte, K.F., Jalleh, R., Wightman, A.J., Douglas, N.J., 1991. The effect of posture on upper airway dimensions in normal subjects and in patients with the sleep apnea/hypopnea syndrome. Am J Respir Crit Care Med 144, 845-847.

Yokoyama, M., Yamanaka, N., Ishii, H., Tamaki, K., Yoshikawa, A., Morita, R., 1996. Evaluation of the pharyngeal airway in obstructive sleep apnea: study by ultrafast MR imaging. Acta Otolaryngol Suppl 523, 242-244.

Yoo, P.B., Durand, D.M., 2005. Effects of selective hypoglossal nerve stimulation on canine upper airway mechanics. J Appl Physiol 99, 937-943.

Younes, M., 2003. Contributions of upper airway mechanics and control mechanisms to severity of obstructive apnea. Am J Respir Crit Care Med 168, 645-658.

Younes, M., 2004. Role of arousals in the pathogenesis of obstructive sleep apnea. Am J Respir Crit Care Med 169, 623-633.

Younes, M., 2008. Role of respiratory control mechanisms in the pathogenesis of obstructive sleep disorders. J Appl Physiol 105, 1389-1405.

Younes, M., Ostrowski, M., Atkar, R., Laprairie, J., Siemens, A., Hanly, P., 2007. Mechanisms of breathing instability in patients with obstructive sleep apnea. J Appl Physiol 103, 1929-1941.

Younes, M., Ostrowski, M., Thompson, W., Leslie, C., Shewchuk, W., 2001. Chemical control stability in patients with obstructive sleep apnea. Am J Respir Crit Care Med 163, 1181-1190.

Young, T., Palta, M., Dempsey, J., Peppard, P.E., Nieto, F.J., Hla, K.M., 2009. Burden of sleep apnea: rationale, design, and major findings of the Wisconsin Sleep Cohort study. WMJ 108, 246-249.

Young, T., Palta, M., Dempsey, J., Skatrud, J., Weber, S., Badr, S., 1993. The occurrence of sleep- disordered breathing among middle-aged adults. N Engl J Med 328, 1230-1235.

Young, T., Skatrud, J., Peppard, P.E., 2004. Risk factors for obstructive sleep apnea in adults. JAMA 291, 2013-2016.

Zaidi, F.N., Meadows, P., Jacobowitz, O., Davidson, T.M., 2013. Tongue anatomy and physiology, the scientific basis for a novel targeted neurostimulation system designed for the treatment of obstructive sleep apnea. Neuromodulation 16, 376-386; discussion 386.

Zamora, N., Llamas, J.M., Cibrian, R., Gandia, J.L., Paredes, V., 2011. Cephalometric measurements from 3D reconstructed images compared with conventional 2D images. Angle Orthod 81, 856-864.

Ziskin, M.C., 1993. Fundamental physics of ultrasound and its propagation in tissue. Radiographics 13, 705-709.

Ziskin, M.C., Thickman, D.I., Goldenberg, N.J., Lapayowker, M.S., Becker, J.M., 1982. The comet tail artifact. J Ultrasound Med 1, 1-7.

Zur, K.B., Mu, L., Sanders, I., 2004. Distribution pattern of the human lingual nerve. Clin Anat 17, 88- 92.

190