Enhanced Electrogastrography Using Transcutaneous Intraluminal Impedance Measurements (TIIM)

University of Calgary PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2016 Enhanced Electrogastrography Using Transcutaneous Intraluminal Impedance Measurements (TIIM)

Poscente, Michael Dennis

Poscente, M. D. (2016). Enhanced Electrogastrography Using Transcutaneous Intraluminal Impedance Measurements (TIIM) (Unpublished master's thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/25899 http://hdl.handle.net/11023/3201 master thesis

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY

Enhanced Electrogastrography Using Transcutaneous Intraluminal Impedance Measurements (TIIM)

Michael Poscente

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF MASTER OF SCIENCE

GRADUATE PROGRAM IN BIOMEDICAL ENGINEERING

CALGARY, ALBERTA

AUGUST, 2016

Abstract Gastric motility and gastric emptying rates have been implicated in the symptoms of functional dyspepsia and gastroparesis. The current gold standard to assess gastric emptying is scintigraphy, which is limited by standardization and radiation concerns. This highlights the need for a novel method of assessing gastric motility.

Transcutaneous Intraluminal Impedance Measurement (TIIM) is a novel method of assessing gastric motility. By measuring the dynamics of a known signal emitted from a battery- powered gastric retentive oscillator within the stomach, gastric motility can be quantified.

In an eight-dog sham comparison study, TIIM was compared to force transducers implanted on the stomach. Two assessment metrics demonstrated statistically significant Pearson correlation coefficients between active TIIM pills and the force transducers (p<0.01, p<0.05), but not when TIIM pills were replaced by sham deactivated pills (p>0.1, p>0.1). A novel portable TIIM receiver was also proposed and tested in the lab to facilitate future ambulatory studies.

iii

Acknowledgement

I would like to thank Dr. Martin Mintchev for his continual support and dedication throughout my entire undergraduate and graduate education. His encouragement, advice, and teaching were invaluable in my own growth and understanding of the field of biomedical engineering, as well as to my development as a researcher. Dr. Mintchev has given me every opportunity to develop and explore my passion for biomedical science, which along with his embodiment of world-class research set me down the road of life-long learning.

I would like to thank my co-supervisor Dr. Orly Yadid-Pecht. Her enthusiasm and drive for innovation motivated me to think outside the scope of my thesis, and opened my eyes to a whole realm of possibilities.

I would like to thank my mentors and research colleagues Dr. Christopher Andrews, Dr. Gregory Muench, and Barbara Smith for their guidance and assistance throughout the research process. Their skills and expertise were essential to the work presented in this Thesis.

My student colleagues Joseph (Gang) Wang, Dobromir Filip, Gavin (Xuexin) Gao, Leticia M.A. Rodriguez, Faustina (Qian) Lu, Robin Wang, Martin Berka, and Thiago Valentin have all been indispensable throughout all of the research I have been involved in. Dobromir and Gavin taught me more in one summer than I had learned in the 2 years previous, and showed me what was possible with hard work and dedication. I am indebted to Joseph Wang, as his interest and involvement in my research gave me someone with whom I could bounce ideas off of, and I am grateful for his encouragement, insights, and knowledge.

I would like to express my gratitude to the technicians and secretaries in both the Electrical and Computer Engineering Department and the Biomedical Engineering Graduate Program at the University of Calgary, for their inspiration, motivation, and assistance to myself and this project in numerous ways. Chris Simon, John Shelley, Warren Flaman, Frank Hickli, Angela Morton, Pauline Cummings, Caron Currie, Ella Lok, Lisa Bensmiller, Lisa Mayer, and Elizabeth Mullaney all made this possible.

Chelsea Ballesteros has been a pillar to me, and has been continually supportive of all my endeavors, research and otherwise. Her patient caring attitude and sharp keen mind helped me through difficulty and reminded me what was important in life.

I would like to thank my entire family, both extended and immediate, for being genuinely interested in both my research and studies, as well as their investment in my future. My sisters Dana and Sophia Poscente, as well as my parents Paul Poscente and Krista Francis deserve a special thanks for their love and support throughout my entire life. The countless hours of homework help and their engagement in my mental and physical well-being has shaped who I am today, for which I am eternally grateful. Lastly I would like to thank my grandfather Dennis Francis, who I will always be grateful to for his intelligent discourse, and making my education possible.

Table of Contents

Abstract ...... ii

Acknowledgement ...... iii

Table of Contents ...... v

List of Tables ...... viii

List of Figures ...... ix

List of Abbreviations ...... xii

Chapter 1: Introduction ...... 1

1.1 Gastric Motility and its Clinical Significance ...... 1

1.1.1 Anatomy of the Stomach ...... 1

1.1.2 Gastric Motility ...... 1

1.1.3 Clinical Significance of Gastric Motility ...... 4

1.2 Disorders Affecting the Stomach ...... 6

1.2.1 Functional Dyspepsia ...... 6

1.2.2 Gastroparesis ...... 12

1.3 Methods to Assess Gastric Motility ...... 19

1.3.1 Gastric Emptying Scintigraphy ...... 19

1.3.2 Stable Isotope Breath Test ...... 20

1.3.3 Fluoroscopy ...... 21

1.3.4 Antroduodenal Manometry ...... 21

1.3.5 Barostat Balloon ...... 22

1.3.6 Electrogastrography ...... 22

1.3.7 Wireless Motility Capsule ...... 23

1.3.8 Magnetic Resonance Imaging (MRI) ...... 24

1.4 Aim of this Thesis ...... 26

Chapter 2: Transcutaneous Intraluminal Impedance Measurements (TIIM) as Enhanced Electrogastrography ...... 27

2.1 Concept ...... 27

2.2 Catheter Based Transcutaneous Intraluminal Impedance Measurements; a Feasibility Study 29

2.3 Gastric Retentive Pill-Based Design ...... 31

2.3.1 Oscillator Design ...... 31

2.4 Signal Acquisition, Conditioning, Digitization, Processing, and Data Logging ...... 38

2.4.1 Signal Acquisition ...... 38

2.4.2 Signal Conditioning ...... 38

2.4.3 Signal Digitization ...... 39

2.4.4 Signal Processing ...... 39

2.5 Development of a portable TIIM receiver ...... 41

2.5.1 Inputs ...... 42

2.5.2 Amplification and filtering ...... 42

2.5.3 Microcontroller and Bluetooth ...... 46

2.5.4 Battery and portability considerations ...... 46

2.6 Laboratory Testing ...... 47

2.6.1 Laboratory Testing of TIIM capsule ...... 47

2.6.2 Laboratory Testing of Portable TIIM Receiver ...... 48

2.7 Animal Testing ...... 49

Chapter 3: Results ...... 55

vii

3.1 Laboratory Testing ...... 55

3.1.1 Laboratory Testing of the TIIM Capsule ...... 55

3.1.2 Laboratory Testing of the Portable TIIM receiver ...... 56

3.2 Animal Testing ...... 57

3.2.1 Gastric Retention ...... 57

3.2.2 Motility Indices, Pearson Correlation Coefficients, and Contractions per Minute 58

Chapter 4: Discussion...... 63

Chapter 5: Conclusion ...... 69

Chapter 6: Scholarly Contributions ...... 70

References ...... 72

Appendices ...... 86

viii

List of Tables

Table 1-1: A comparison of Functional Dyspepsia and Gastroparesis ...... 19 Table 1-2: Summary of available technologies to assess gastric motility and emptying ...... 25 Table 3-1: Averaged Pearson correlation coefficients (PCCS) of the one-minute gastric motility indices (GMIs) per state per capsule type...... 59 Table 3-2: Averaged cycles per minute (CPM) of the raw force transducer (FT) and cutaneous recordings per state per capsule type...... 60 Table 3-3: Statistical comparison between the dominant frequency peaks of the FT and TIIM/Sham recordings using paired Students t-test. *(p < 0.05)...... 60

List of Figures

Figure 1-1: The Anatomy of the Stomach. The hollow lumen is surrounded by three layers of muscle; oblique, circular, and longitudinal, all of which are encased within the serosa. Orientation of muscle fibers within each layer is shown with a reference gastric axis imposed. Ingested food enters the stomach from the esophagus via the lower esophageal sphincter. The fundus accommodates food and propels it towards the antrum, where it is crushed and mixed. Ingested content is held in the stomach by the pyloric sphincter, which only allows chyme to pass at certain sizes...... 3 Figure 1-2: The diagnosis process of functional dyspepsia...... 9 Figure 1-3: The diagnosis process of gastroparesis...... 14

Figure 2-1: A simplified electrical model of the stomach with a TIIM capsule, VTIIM, injecting

current into the tissue. RTissue and RSkin are constant when compared to RMuscle, and changes to

the impedance of RMuscle will affect the voltage measured at RSkin. Without the TIIM capsule,

the inherent electrical activity denoted by VAC is not strong enough to create detectable

changes in the voltage across RSkin...... 29 Figure 2-2: Overall catheter-based experimental setup. 1. Stomach; 2.1. Distal force transducer; 2.2. Proximal force transducer; 3. Connecting wire from force transducers; 4. Transoral catheter; 5. Custom bridge amplifier; 6. ECG electrodes; 7. Receiving wires from ECG electrodes; 8. External TIIM oscillator with isolated ground; 9. Isolated bioelectric amplifier; 10. Analog-to-digital converter; 11. Real time data acquisition software...... 30 Figure 2-3: The oscillator circuit, printed circuit board layout, and TIIM capsule body. Dimensions of the circuit board were 10mm x 6mm, which allowed it to fit inside the capsule body that had dimensions 19mm x 10mm...... 34 Figure 2-4: Breakdown of the TIIM gastric retentive pill. 1. Oscillator circuit; 2. Capsule body; 3. Assembled capsule; 4. Superabsorbent granules; 5. Capsule and granules inside a liquid- permeable mesh; 6. Dissolvable pill containing the mesh enclosed capsule; 7. TIIM gastric retentive pill; 8. Pill expanded in water; 9. Test dish. The oscillator circuit was first sealed into the assembled capsule, and added to superabsorbent granules held in a liquid-permeable

mesh. This was inserted into a dissolvable pill, which disintegrated in the stomach, allowing the granules to swell. A mm scale is provided on the left of the image...... 37 Figure 2-5: The analog amplification and conditioning circuit implemented in the portable TIIM receiver. This circuit is implemented twice to provide two TIIM channels on the prototype receiver. Variable gain and filtering parameters can be set using the DIP switch between the two stages. The offset of the output voltage can be adjusted using the 20kΩ potentiometer. In this configuration the gain is set to 1000X with no filtering option selected...... 43 Figure 2-6: DIP Switch Reference Chart ...... 45 Figure 2-7: The TIIM receiver and its components...... 47 Figure 2-8: The experimental setup to assess the portable TIIM receiver, as well as the circuit diagram of the attenuation circuit...... 49 Figure 2-9: The serosal view of the stomach showing the placement of the force transducers (A), and the internal view of the stomach showing the position of the TIIM capsule and gastric retentive mesh...... 51 Figure 2-10: Overall gastric retentive pill-based experimental setup. 1. Stomach; 2. Gastric retentive mesh; 3. TIIM capsule; 4. Implanted force transducers; 5. Connecting wire from force transducers; 6. Custom bridge amplifier; 7. Cutaneous electrodes; 8. Receiving wires from ECG electrodes; 9. Isolated bioelectric amplifier; 10. Analog-to-digital converter; 11. Real-time data acquisition software...... 52 Figure 2-11: The overall experimental procedure...... 53 Figure 3-1: An oscilloscope reading from cutaneous electrodes attached to tripe to verify the frequency of the TIIM oscillator inside (A). Palpations were visible as amplitude modulations (B)...... 55 Figure 3-2: Combined plot of the pre-attenuated signal, portable receiver recorded signal, and their respective one-minute motility indices. The input signal was taken from a 30-minute recording obtained during animal testing...... 56 Figure 3-3: An oscilloscope reading from the gastric serosa prior to the force transducer implantation verified the presence of an activated TIIM pill (A). The sham pills did not demonstrate any signal (B)...... 57

Figure 3-4: Combined plot of the raw signals and the one-minute motility indices for an active pill in the baseline state (0 - 1800) and post-neostigmine administration (1800 – 3600). A thick vertical line denotes the administration of neostigmine...... 61 Figure 3-5: Combined plot of the raw signals and the one-minute motility indices for an inactive pill in the baseline state (0 – 1800) and post-neostigmine administration (1800 – 3600). A thick vertical line denotes the administration of neostigmine...... 62 Figure 4-1: Proposed technique for measuring the impedance of the liver. In this case the internal transducer may be acoustic or electrical, depending on the desired results...... 68

xii

List of Abbreviations AC Alternating Current AWG American Wire Gauge CMOS Complementary Metal-Oxide Semiconductor CPM Contractions Per Minute DAQ Digital Acquisition DC Direct Current DIP Dual Inline Package ECG Electrocardiograph EGG Electrogastrograph FT Force Transducer GI Gastrointestinal GMI Gastric Motility Index H2RA Histamine Type 2 Receptor Antagonist ICC Interstial Cells of Cajal IEEE Institute of Electrical and Electronics Engineers MMC Migrating Motor Complex MRI Magnetic Resonance Imaging Op-amp Operational Amplifier PCC Pearson Correlation Coefficient PCMCIA Personal Computer Memory Card International Association PEI Polyetherimide PPI Proton Pump Inhibitor PTFE Polytetrafluoroethylene PVA Polyvinyl Alcohol R U Relative Units SSRI Serotonin Reuptake Inhibitors TCA Tricyclic Antidepressants TIIM Transcutaneous Intraluminal Impedance Measurement

Chapter 1: Introduction 1.1 Gastric Motility and its Clinical Significance 1.1.1 Anatomy of the Stomach The stomach is a foregut derived organ [1] that is responsible for mechanically accommodating, grinding, and mixing food, as well as chemically secreting acid and enzymes to assist in the breakdown of food [2]. The human stomach can be divided into two main sections based on mechanical function and location; the proximal fundus and the distal antrum. Ingested food travels down the esophagus, and passes through the lower esophageal sphincter where it enters the fundus. The fundus acts as a reservoir that relaxes and extends when food enters the stomach and maintains pressure to keep food in the main digesting antral part of the stomach. The antrum of the stomach is responsible for high intensity contractions that grind and mix ingested food, as well as provide pressure across the pyloric sphincter to propel chyme into the small intestine at a controlled rate. Ingested content is driven towards the pyloric sphincter by propulsive peristaltic contractions, but if it is still too large it travels back towards the fundus to be crushed again via retropulsive peristaltic contractions, which further help to crush and mix food into chyme [2].

1.1.2 Gastric Motility The stomach is made up of smooth muscle, and has no extracellular matrix like bone associated with skeletal muscle for the opposing muscle groups to act against. Instead opposing muscle groups attach directly to each other, and can be classified based on orientation to the organ and discrete layer the muscle group occupies. The inner most layer is the oblique muscle, and is unique to the stomach. The oblique muscle adds strength and thickness to the stomach, particularly around the pylorus and antrum [3]. The muscle fibers in this layer run obliquely to the organ, giving them their name. This layer exhibits powerful motion, that facilitates effective crushing and mixing. The middle layer is the circular muscle, lying perpendicular to the gastric axis in a ring around the organ. Contractions of the circular muscle cause a decrease in the diameter of the stomach at the location of the contraction. This layer is continuous throughout

the stomach, and forms the pyloric sphincter. The most superficial muscle layer is the longitudinal muscle, which lies parallel to the gastric axis. The muscular layers and regions of the stomach are shown in Figure 1-1. Contractions of circular muscles require the longitudinal muscle to relax, and vice versa, whereas the oblique muscles can work in supportive or opposing motion to either muscle group depending on the state of digestion and their location. When the longitudinal muscle contracts it decreases the length of the stomach, and pulls the non-contracting circular muscle into a larger diameter [3]. By coordinating these contractions spatially and temporally the organ is able to achieve complex tasks, such as moving a bolus along its length, mixing stomach contents with digestive enzymes and chemicals, or mechanical digestion of the bolus into chyme.

The peristaltic motions of the smooth muscle in the stomach can vary in intensity and rate, and as such have very complex control mechanisms and structures that incorporate the faster autonomic nervous system and the slower endocrine system [4, 5]. Sympathetic innervation comes from the gastric plexus of the celiac ganglion, coming from the thoracic region of the spine, and is associated with providing contraction and inhibiting peristalsis. Parasympathetic innervation comes from the vagus nerve, also known as cranial nerve 10, and stimulates gastric secretion of acid from parietal cells and promotes peristalsis. Mechano- receptor afferent nerves in the stomach detect stretching as food is ingested, which triggers parasympathetic gastric secretion and strong rhythmic contractions in the antrum, as well as the release of the paracrine hormone gastrin. Chemo-receptor nerves in the small intestine detect the presence of nutrients, in particular fat, and trigger alternating patterns of propulsive and retropulsive peristalsis to better break down food. This increases the food’s surface area, as well as keeps food in the stomach longer [5, 6]. Since we hydrate and receive nutrients along the same pathway it is interesting to consider the ingestion of water. Ingesting water triggers the mechano-receptors, and as such triggers contractions in the antrum. As water is emptied from the stomach the chemo and osmo-receptors in the small intestine are not triggered by nutrients or fat, and the stomach is elicited to rapidly empty, allowing quick and efficient uptake for effective rehydration.

Within the endocrine control mechanisms of the stomach there are a variety of hormones that have varying regional effects. Two hormones, gastrin and cholecystokinin have an excitatory effect in the antrum, enhancing contractions, while having an inhibitory effect in the fundus, causing it to relax to better accommodate food [7]. While both of these functions are necessary for regular digestion, their distinct differences in regional effect of the same organ demonstrates that patterns of gastric motility are the result of the integration of a large number of inhibitory and stimulating signals from both nervous and endocrine channels. Dopamine has an inhibitory effect on gastric motility, and has been shown to significantly delay gastric emptying. Dopamine inhibitor drugs have demonstrated promise in clinical treatments of delayed gastric emptying [8].

Unlike in the heart, contractile activity has been shown to dissociate from the electrical activity in the stomach [9], and the electrical waves from the stomach are much weaker in amplitude than those from the depolarization of cardiac muscle in the heart. In the heart, electrical activity is conducted directly between muscle cells to produce a uniform contraction, whereas in the stomach, cells located between the muscle layers called the Interstitial Cells of Cajal (ICC) are responsible for providing a slow electrical signal that triggers contraction of the smooth muscle [10, 11]. Action potentials of smooth muscle cells do not appear to propagate over large distances, and thus coordinated movement is almost entirely dependent on the spatiotemporal patterns of the slow waves generated by pacemaker networks of the ICC. These ICC cells mediate input from efferent motor neurons, and help coordinate regular peristalsis at a rate of approximately 3 waves per minute during peak digestion in humans.

During interdigestive periods the stomach is still active, and goes through a regular cycle called the migrating motor complex (MMC), which is triggered by the hormone motilin [12]. The MMC can be thought of as housekeeping for the gastrointestinal tract, as it works to remove anything undigestible from the stomach and GI tract, and promotes gastric secretions to help eliminate bacterial growth. The MMC cycle occurs every 1.5 to 2 hours, depending on how recently the last digestive state occurred, and is characterized by 35 to 40 minutes of contractions that increase in frequency and intensity, then a period of quiescence that can last 45 to 60 minutes [12]. During the MMC the pyloric sphincter is relaxed, allowing undigestible substances or similar to pass through the remaining GI tract. The MMC is responsible for the growling and gurgling sounds from the GI tract between digestive periods, and in the event of food intake the MMC will be replaced by regular digestive contractions.

1.1.3 Clinical Significance of Gastric Motility Due to the complexity of the control mechanisms for gastric motility, and despite built in redundancy there is the possibility for problems to arise, but as of yet we still have poor understanding of the diseases that can compromise gastric motility [13, 14]. One of the most

basic scenarios is spinal cord damage. Depending on the location of the damage, gastric ileus (paralysis) may occur, in which the stomach is temporarily paralyzed [15]. If the vagus nerve is not damaged by the injury it will still promote secretion of acid in the stomach, which if left undisturbed can cause ulcers and damage to the stomach lining until regular function can be restored [16].

Conversely, if damage occurs to the vagus nerve, which can result from diabetes-related neuropathy, gastroparesis can be induced. This results in slower and less intense contractions [17], and as a result can negatively affect nutrient uptake, inhibiting proper glucose management [18]. When food is not being mechanically or chemically digested to a sufficient degree in the stomach to facilitate absorption in the small intestine, patients can experience nausea, early satiety, and abdominal pain. A further complication can be the fermentation of food in the stomach, a result of bacterial overgrowth in the absence of sufficient gastric acid secretion [19].

Damage to the ICC cells can cause dysrhythmias to occur, and may contribute to symptoms associated with gastroparesis, functional dyspepsia, and irritable bowel syndrome in the large intestine [20]. The ICC cells need to be in a continuous network to provide effective contractions for peristalsis to occur [11], and damage can cause spasms in the muscle tissue, similar to damage to conductive muscle fibers in the heart.

The stomach is a largely underappreciated organ that plays a critical role in nutrition, health, and general well-being. As imaging and diagnostic techniques develop to give insight into the disorders and diseases that affect motility, we may be able to better understand the complex etiology behind gastric motility disorders.

1.2 Disorders Affecting the Stomach 1.2.1 Functional Dyspepsia 1.2.1.1 Overview of Functional Dyspepsia Dyspepsia refers to a broad range of symptoms affecting the gastroduodenal region, ranging from early satiety to pain in the epigastric area, and is also known as indigestion. Functional dyspepsia is the presence of dyspeptic symptoms without any indication of an organic explanation. It is estimated to have a global prevalence between 5 and 15%, and was assessed to have cost the United States $18 billion in 2009 [21]. Due to the ambiguous nature of functional dyspepsia, diagnosis involves exclusion of other possible causes while meeting the Rome III criteria for at least 3 months with symptoms onset 6 months prior to diagnosis [22]. During this period upper endoscopy will be performed to eliminate the possibility of mechanical obstruction. The Rome III criteria for functional dyspepsia are summarized below.

Rome III criteria for Functional Dyspepsia [23]

Diagnostic criteria: Must include one of more of the following with no evidence of structural disease o Bothersome postprandial fullness o Early satiation o Epigastric pain o Epigastric burning A. Postprandial distress syndrome o Bothersome postprandial fullness after ordinary meals o Early satiation that prevents finishing a meal Supportive criteria o Upper abdominal bloating o Epigastric pain syndrome may coexist B. Epigastric Pain Syndrome o Pain or burning in the area of the epigastrium of at least moderate severity at least once per week o Pain is intermittent o Not generalized or localized to other abdominal or chest regions o Not relieved by defecation or passage of flatus o Not fulfilling criteria for biliary pain Supportive criteria o Pain may be of burning quality o Pain is commonly induced or relieved by ingestion of a meal o Postprandial distress syndrome may coexist

Similar symptoms can be caused by a variety of other issues, including peptic ulcers, gastroesophageal cancer, gastroparesis, pancreatitis, parasites, and Crohn’s disease. Tests to distinguish between functional dyspepsia and gastric cancer are only performed in the presence of alarm symptoms [24]. Although not included in the Rome III criteria, patients with functional dyspepsia also often report heartburn [25], making it difficult to rule out gastroesophageal reflux disease, and often an empirical trial period of proton pump inhibiting drugs to suppress gastric acid secretion is attempted before further investigation. Functional dyspepsia can also coexist with irritable bowel syndrome, further complicating the symptoms presented by patients [22].

Up to 25% of functional dyspepsia patients demonstrate some degree of delayed gastric emptying, a symptom also associated with gastroparesis. In most cases of functional dyspepsia, symptoms do not consistently predict the underlying pathophysiology, making the Rome III criteria inadequate for guiding therapy [25]. Functional dyspepsia has been shown to cause anxiety about eating, and can lead to eating disorders and depression [26]. Diagnosis of functional dyspepsia is typically a process of elimination, in which patients are exposed to potentially harmful therapies and expensive, time consuming examinations, which further increases the stress and frustration patients encounter [21]. These tests can involve abdominal ultrasounds, endoscopies, and scintigraphic evaluation to assess gastric emptying. Only once these tests prove to be inconclusive can a definitive diagnosis of functional dyspepsia be made. A flow chart included in Figure 1-2 demonstrates a sample diagnosis process.

Figure 1-2: The diagnosis process of functional dyspepsia.

1.2.1.2 Treatment and Management Options There is no medication that is currently approved to treat functional dyspepsia in the US, Canada, or the European Union [21]. There are treatment options, however it is difficult for physicians to suggest which treatment will be most effective as the underlying causes of functional dyspepsia are often not understood due to lack of availability of testing [27, 28].

Helicobacter pylori Eradication In some patients a previous gastrointestinal infection will lead to functional dyspepsia. Helicobacter pylori is a bacterium that causes one of the most common gastric infections, and can cause gastric inflammation. The involvement of antibiotics in treatment for H. pylori, and

general concern regarding widening antibiotic resistance make this therapy difficult to recommend without first positively identifying an infection [29-31].

Proton Pump Inhibitors Proton Pump Inhibitors (PPI) were first demonstrated to be successful in the treatment of gastroesophageal reflux disease and peptic ulcers, and have become widely used in the treatment of dyspepsia. PPIs work by inhibiting the gastric proton pump of the parietal cells found in the luminal walls of the stomach. This proton pump is directly responsible for secreting H+ ions, and PPIs are one of the most potent acid inhibitors available. PPIs are often given empirically, however high doses of PPI were shown to be ineffective in predicting functional dyspepsia patient symptom outcomes [32, 33]. When patient groups were limited to “dysmotility-like” dyspepsia symptoms there was no advantage found to using PPIs. The most beneficial effect of PPIs in treating functional dyspepsia may be through the management of coexisting reflux symptoms [34]. The Food and Drug Administration recommends limited use of PPIs, and as such patients that do not respond to PPIs should discontinue use after 4-8 weeks [35].

Prokinetic Agents Prokinetic agents are a class of drugs that enhance gastrointestinal motility, by increasing the frequency and strength of contractions, particularly in the small and large intestines. Prokinetic agents increase acetylcholine concentrations by a variety of pathways, stimulating gastrointestinal motility and improving coordination between the stomach and duodenum [36]. Prokinetic agents have also been demonstrated to increase colonic motility. Despite the seemingly promising effects of prokinetic agents there was limited improvement of symptom management in randomized controlled trials for patients with functional dyspepsia [32] .

Histamine Type-2 Receptor Antagonists Histamine Type-2 Receptor Antagonists (H2RA) block histamine receptors on the parietal cells in the stomach. This decreases the production of stomach acid, and can be used when treating dyspepsia, peptic ulcer disease, and gastroesophageal reflux disease [36]. H2RA therapy has been evaluated with respect to individual dyspepsia symptoms. Postprandial fullness and epigastric pain were improved; however, the results were not significant for any other symptoms [32]. With functional dyspepsia there is a possibility for associated gastroesophageal reflux disease, and a reduction in heartburn symptoms may also explain the results of these trials. H2RA therapy is less effective than proton pump inhibitors, which are more widely used [32].

Tricyclic Antidepressants Tricyclic Antidepressants (TCA) are used to treat chronic pain, and have been shown to be effective in treating irritable bowel syndrome. TCAs do not affect the motor function of the gastrointestinal tract, and do not improve satiety. TCAs have been shown to improve symptoms of functional dyspepsia, particularly in patients not responding to PPI or prokinetic therapies, however the studies were very small, and it was not clear whether the drugs were reducing pain in the nervous system or impacting anxiety or depression that may be present in functional dyspepsia patients [21, 37].

Serotonin Reuptake Inhibitors Serotonin reuptake inhibitors (SSRI) are also used to treat irritable bowel syndrome, which can coexist with functional dyspepsia. SSRIs may improve pain in some patients [38], however limited research has been done to assess the clinical use of SSRIs in functional dyspepsia patients, and the research done has found little to no effect in reducing patient’s symptoms when compared to a placebo [39].

Diet Modification Patients often modify their dietary habits in order to consume foods that are less likely to induce symptoms in their particular case. Consumption of fat has been implicated as symptom inducing, however, due to the variety of the pathologies each patient is likely to respond differently. Diet modification is often ignored as a treatment option because of lack of clinical data, however consuming smaller meals with reduced fat content offers potential as a symptom management option [27].

The broad array of treatment options for functional dyspepsia highlight the diversity of potential etiologies for the disorder. Abnormalities in gastric motility are present in 25% of patients with functional dyspepsia, however further studies are required to assess the relationship between gastric motility and the symptoms presented by patients [27].

1.2.2 Gastroparesis Gastroparesis does not have symptom-based criteria like the Rome III criteria to meet, but is still diagnosed through an exclusionary process. The reason for this exemption is that although gastroparesis has a gastric motility reducing effect, there are several organic etiologies that can result in delayed gastric emptying. Gastroparesis is defined as delayed gastric emptying in the absence of mechanical obstruction, which excludes gastric cancers. The symptoms of gastroparesis are very similar to functional dyspepsia, and include early satiety, postprandial fullness, nausea, vomiting, and upper abdominal pain or discomfort [40, 41]. There are three primary classifications of gastroparesis; diabetic, post-surgical, and idiopathic. When ingested food is allowed to sit in the stomach for an extended period of time as in the case of delayed gastric emptying, it is possible for fermentation of the food to occur, which has dangerous implications for the individual affected [42]. Weight loss, malnutrition, and dehydration are all possibilities in extreme cases of gastroparesis. These cases can involve hospitalization in order to rehydrate a patient, and chronic care can include enteral or parenteral nutritional maintenance [43, 44]. The diagnosis process is outlined in Figure 1-3.

1.2.2.1 Diabetic Gastroparesis Complications with long-term management of diabetes are the leading causes of gastroparesis in the United States. Conservative estimates from leading research centers place the prevalence of gastroparetic symptoms in the range of 40 % of Type 1 diabetics, and up to 20% of Type 2 diabetics [45]. This statistic represents 4 million diabetic adult Americans suffering from the indicators of gastroparesis, even if it has not been confirmed by a physician. Delayed gastric emptying associated with any cause can inhibit a diabetic patient’s ability to manage and control their nutrition and caloric intake, further exacerbating the factors that lead to the development of diabetes [46]. Diabetic gastroparesis can be a result of neuropathy, particularly damage to the vagal nerve that is responsible for initiating gastric contractions. With diabetic patients there is little correlation between symptoms and the rate of gastric emptying [47, 48], only postprandial fullness is indicative of the speed at which ingested solid content is emptied. This suggests that there are likely other factors that influence the presence and severity of symptoms [49]. For diabetic patients, a priority for treatment should be to strive for optimal glycemic control, as hyperglycemia has an inhibitory effect on gastric emptying [50]. High blood sugar can adversely affect response to therapies, particularly the stimulatory effect of prokinetic drugs like erythromycin [51]. With diabetic patients, aggressive glucose monitoring is important because of the spiraling effect changes in blood glucose can have on a patient’s ability to orally control their blood sugar.

1.2.2.2 Post-surgical and Injury Related Gastroparesis As in the case of diabetic gastroparesis, surgical damage to either branch of the vagal nerve or the support structures around it will inhibit the control of the peristaltic motion of the stomach and small and large intestines [44]. Surgical operations that involve the stomach or the region of the stomach can cause temporary gastric ileus, which can be hazardous to the health of a patient as stationary gastric secretions can damage the mucosal lining of the lumen. This can lead to irritations to the tissue of the stomach, and ultimately can result in ulcers [52]. In

some cases, depending on the extent of the damage this paralysis of the stomach can be temporary, however in more dramatic procedures such as vagotomy in which the vagal nerve is severed gastric motility function will strictly be the result of the inherent feedback mechanisms of the stomach [53].

1.2.2.3 Idiopathic Gastroparesis As with functional dyspepsia there are cases of gastroparesis that are presented without any identifiable underlying cause. In this case the cause is labelled as idiopathic gastroparesis, and is estimated to affect 36 % of gastroparesis patients [40]. Diagnosis of idiopathic gastroparesis is completed in a similar fashion to that of functional dyspepsia, with similar effects on patient morale. Modern understanding of the pathophysiology of idiopathic gastroparesis is incomplete, particularly in the mechanisms that lead to the symptoms presented by patients [54].

Figure 1-3: The diagnosis process of gastroparesis.

1.2.2.4 Treatment and Management Options Effective management of gastroparesis requires a detailed understanding of the pathogenesis of the disorder, particularly as to whether it is a result of a deficit in muscular or

neurological function [40]. Regional irregularity in the stomach has been associated with different symptoms, and as such different therapies have been developed that address specific complaints. Investigation into the severity of the disease can also indicate which patients are likely to benefit from different treatment options. The main priorities of therapy are to address fluid and nutritional deficiency, reduce symptoms, and identify and correct the underlying cause. Management of gastroparesis typically involves modification of a patient’s diet, as well as stimulating gastric activity [55, 56].

Dietary and Nutritional Recommendations Unlike in functional dyspepsia, changes to a patient’s diet are well documented as an effective technique to reduce gastroparesis symptoms, particularly in patients in which milder forms of gastroparesis are present. Dietary interventions can be made considering the properties of different types of food, and how they interact with the stomach and small intestine. It is possible that intolerances to different foods can adversely affect an individual’s gastric motility, which should be reduced or eliminated in order to minimize their effect [57]. Consuming smaller meals more often is also recommended to reduce distension, as well as consuming more fluids and avoiding foods that are difficult to chew. Fatty foods should be restricted [56], as lipids delay emptying, and high fiber foods or treatment for constipation should be avoided as these can promote bezoar formation [58, 59]. Patients may need to supplement their diet with multivitamins, or even liquid meals if caloric needs aren’t being met. In the case of diabetic patients, meal timing and composition are particularly critical as they are more prone to dramatic variations in blood glucose levels, which negatively affect gastric motility [60].

Prokinetic Drugs Prokinetic medications enhance contractility of the GI tract, and are most effective at combatting the symptoms of gastroparesis when implemented in conjunction with dopamine receptor antagonists. By increasing acetylcholine in the stomach and thereby stimulating gastric

motility several symptoms associated with gastroparesis can be reduced, namely postprandial fullness, early satiety, and in some cases nausea and vomiting [61]. A regimen of prokinetic drugs and dopamine receptor antagonists is a common first line of treatment for gastroparesis [62-64].

Dopamine Receptor Antagonists Dopamine acts as an inhibitor of gastric activity [8], and so dopamine receptor antagonists counteract the effects of dopamine on gastric emptying. Patients admitted to the hospital for gastroparesis will often receive dopamine receptor antagonist drugs intravenously, along with electrolytes for rehydration. Dopamine further acts as an antiemetic agent on the brainstem, which further reduces symptoms [65].

Motilin Receptor Agonists Motilin elicits antroduodenal contractions, and a number of antibiotics act as motilin receptor agonists to promote gastric motility [51]. There has been a push in pharmaceutical investigation to establish motilin agonists that exhibit prokinetic capabilities. A variety of motilin receptor agonists have been developed with very specific criteria for effective use, such as for use in diabetic patients with or without nausea [66].

Antiemetic Medication Antiemetic medications act to reduce vomiting, and often antiemetic drugs without the potential for prokinetic effects are commonly used to reduce nausea and vomiting associated with gastroparesis [65]. The pathways by which antiemetic drugs operate are diverse, and with a wide variety of drugs available the prescribing clinician can take into account different side effects in making a selection [67].

Psychological Support Relentless nausea and vomiting can result in damaging psychological consequences, particularly when associated with pain [68]. Both gastroparesis and functional dyspepsia have a negative effect on a patient’s quality of life, as in both cases it is possible that the disease will be chronic, and despite doctor’s efforts all that can be done is to manage symptoms. Simple practices such as relaxation techniques and distraction can help promote a sense of well-being for the patient [18, 69].

Enteral Nutritional Support Enteral nutritional support is necessary when the caloric and fluid needs of a patient are not being met orally. Enteral support is used sparingly because of its invasive nature, and access points come with advantages and complications for the type of support that is trying to be accomplished. Even more invasive is parenteral nutritional support, which for a patient with chronic gastroparesis is only appropriate in emergency cases [70-72].

Endoscopic Therapy Botulinum toxin delivered endoscopically to the pylorus will promote muscle relaxation, and is used in cases of functional gastric obstruction in which the pylorus exhibits prolonged phasic activity. Higher doses have been associated with reductions in nausea and vomiting, and had a similar effect across all three primary types of gastroparesis [73]. With botulinum injections there is a limited effective time, which was reported as averaging 5 weeks, which must be considered in conjunction with the toxicity of the treatment when suggested to a patient [74], and as botox has not been effective in randomized controlled trials is not recommended as an effective treatment.

Gastric Electrical Stimulation Gastric pacemakers that deliver high energy stimuli to the stomach show promise in promoting gastric emptying and reducing symptoms associated with gastroparesis [75]. Pacing electrodes sewn to the gastric serosa are required to deliver large currents, supplied by an

internal pulse generator implanted in a subcutaneous pocket. Implantation of a gastric electrical stimulator has helped many individuals reduce their reliance on enteral feeding [76], and many were able to discontinue use altogether. This technology is still under constant development, and requires follow-up surgery for removal in the event of complications [77]. Gastric electrical stimulation reduced vomiting frequency in patients [78], and limited the need for pharmaceutical intervention. The most common complications of this technology include infection, wire breakage, dislodgement, and intestinal obstruction [79, 80]. With implanted gastric electric stimulators patients should avoid metal screening security devices, as well as magnetic resonance imaging.

Other Surgical Options Other surgical options are related to drainage of the stomach, such as pyloroplasty and partial or total gastric resections to bypass the stomach. These treatments offer limited symptom management, and are only recommended in the most extreme cases, in which patients do not respond to any of the above therapies [81].

In most cases of gastroparesis symptoms can be difficult to control, which directly affects the quality of life of the affected patient [14]. Adding to that, the exclusionary tests involved in properly identifying gastroparesis can be stressful and the costs of the tests themselves and time off work further burden patients [82]. Frustration is common [83], and reviews of common gastric motility disorders all indicate the need for a deeper understanding of the etiology presented in order to best provide the most effective treatment options to a patient [84]. A brief summary of the two diseases is presented in Table 1-1.

Table 1-1: A comparison of Functional Dyspepsia and Gastroparesis

Disorder Functional Dyspepsia Gastroparesis

Prevalence 5 - 20 % 1 - 2 %

Upper endoscopy, gastric motility Upper endoscopy, gastric motility Diagnosis assessment, PPI treatment, elimination of assessment, elimination of other causes other causes

Motility enhancing drugs, gastric Acid suppression, motility enhancing electrical stimulator, endoscopic therapy, Treatment Options drugs, anti depressants, diet modification nutritional support, diet modification, other surgical options

1.3 Methods to Assess Gastric Motility 1.3.1 Gastric Emptying Scintigraphy Gastric empting scintigraphy was first tested in canines in 1969 by the surgeon Daintree Johnson FRCS [85]. Since then gastric emptying scintigraphy has become the gold standard and most widely used test by which gastric emptying is assessed. This is connected to the fact that required equipment is typically available at radiology centers, and the test itself is minimally- invasive in nature [9, 86]. The most widely adopted method of scintigraphic assessment involves the patient ingesting a radiolabeled meal over the course of 10 minutes, following which gamma cameras acquire images immediately after, 1, 2, and 4 hours after the meal [87]. The clinical use of scintigraphy as a clinical diagnostic tool has been criticized for a number of factors, including lack of meal standardization, variable patient positioning, and even the timing of the images [88]. Furthermore, the exposure to radiation limits the technology’s use in children and pregnant women [89]. While being allowed to move around when not being imaged, patients are required to stay at the imaging facility for the duration of the test. An added issue is the possibility for the radiolabel to dissociate from the meal, leading to overestimation of the emptying rate as the radiolabel leaves the stomach via liquid gastric secretions rather than the intended food [90]. Using scintigraphy, it is usually clear when delayed emptying is present, as determined by the remaining percentage of food within the

stomach at the given time points. Accelerated emptying is more difficult to detect however due to the limited images taken early on [91].

Within healthy individuals there can be significant variation (12-15%) in gastric emptying rates [92], further complicating diagnosis. Studies have demonstrated the need for longer term assessment of gastric emptying [85], as tests lasting less than two hours led to inaccurate assumptions about the gastric emptying rate.

Gastric emptying and motility are closely related [93], and any new reliable gastric motility diagnostic technique can be expected to be able to assess, directly or indirectly, gastric emptying. Scintigraphy can only be used to diagnose gastroparesis after excluding mechanical disease, which is usually done with an upper endoscopy. With scintigraphy there is high interpatient variability, and changes in body position during the test can result in slower emptying times, thus making the test unreliable [94].

1.3.2 Stable Isotope Breath Test The C13 breath test is a modern extension of gastric emptying scintigraphy [95]. Solid or liquid test meals are marked with either C13 or C14, depending on the requirements of the study. After the meal is triturated from the stomach it is digested by the small intestine and

13 13 metabolized by the liver [96]. The metabolized C is detectable in the form of CO2 expired via

13 the lungs. Any change to baseline CO2 on the breath is assumed to be a result of the ingested labeled test meal [97]. To perform a breath test the patient exhales via a straw into a sealable container, which is labeled with the time. Samples are acquired before the test, and at 45, 90, 120, and 180 minutes after ingestion. Once collected samples can be sent to a laboratory to

13 process the CO2 content, making this test available to be performed in remote locations. Reducing the radiation hazard makes this technique suitable for pregnant women and children, and has been proven to be as accurate as scintigraphy [98], earning its endorsement by the

American and European Neurogastroenterology and Motility Societies [99]. This technique is only pseudo-continuous, and does not demonstrate slight variations in emptying time. Stable isotope breath tests are also unable to demonstrate gastric tone, and as such are unable to assess potential neuropathy or myopathy. Relying on metabolism and excretion via the lungs limits us in patients dealing with other diseases related to the intestinal mucosa, liver, and lungs [99].

1.3.3 Fluoroscopy Fluoroscopy utilizes X-rays to image different tissues and structures in the body. Since the gastrointestinal tract is composed of soft tissue, a barium contrast agent has to be consumed in order to make sense of the images. Barium enhanced fluoroscopy scans can attain high temporal resolution, and can detect obstruction and abnormalities [100]. The drawbacks of fluoroscopy are that it doesn’t demonstrate normal meals, as the barium is in a liquid suspension, and delivers a higher dose of radiation than scintigraphy, or even other regular X- rays [101]. Typically, fluoroscopy is more appropriately used for dysphasia and other problems associated with the pharynx and esophagus that require higher spatial-temporal resolution.

Fluoroscopy can demonstrate peristalsis and rate of flow, and small areas can be looked at carefully to identify abnormality and as such is typically better employed to identify ulcers or small scale irritations [100]. Barium is made to coat the stomach to look at the mucosal folds, which may radiate towards ulcers or other inflammation. Fluoroscopy is also ideal for patients that have undergone GI resection. As such fluoroscopy is best suited to being used to eliminate other possibilities of disease more so than as a diagnosis of a functional GI disorder [102].

1.3.4 Antroduodenal Manometry Antroduodenal manometry utilizes either a water perfusion or solid state catheter with pressure transducers placed every 1 cm positioned into the vicinity of the pylorus and antrum

[85, 103]. Following placement of the catheter a meal is consumed, and motility indices can be calculated from the detected pressure patterns in the antrum of the stomach. Reduced motility indices are significantly correlated with delayed gastric emptying [104]. The applicability of antroduodenal manometry is limited due to difficulty interpreting the test and the potential for migration of the catheter [105]. This test can be performed in stationary or ambulatory conditions, and is most effective for detecting cases of pseudo-obstruction, which can be difficult to detect using radiological methods. Manometric results can identify myopathy and neuropathy, assisting physicians to guide the course of treatment [106].

Highly portable solid state catheters with excellent resolution can be used in 24-hour ambulatory monitoring [104]. Solid state catheters are more applicable than water perfused catheters that require expensive and bulky recording equipment, however both are susceptible to physiological and movement artifacts such as swallowing, walking, or respiration [107]. The test is quite invasive, and relies on the presence of a skilled technician.

1.3.5 Barostat Balloon Barostat balloons are the gold standard of assessment of tone of hollow organs, and are a well validated technique [108]. A barostat balloon is inserted into the lumen of the stomach, where it is inflated with water. Any changes to the volume of the balloon demonstrate changes in gastric tone, however this technique has been criticized for the potential of the balloon to induce reflexive relaxation and distension. These tests can last for multiple hours, and the intubation and inflation of the balloon can be stressful and uncomfortable for patients [109]. Barostat balloons are more commonly used in assessing anorectal tone [110].

1.3.6 Electrogastrography Monitoring the natural electrical activity of the stomach via electrogastrography (EGG) has also been proposed as a method to detect gastric motility [111]. The electrical rhythm of

the gastric slow waves has been shown to dissociate from the frequency of gastric contractions, and there is still much debate about the optimal placement of electrodes [112]. Recent research with gastric electrical mapping in live human stomachs has shown slow wave activity to be very complex, often with multiple wave fronts propagating in the stomach simultaneously [113]. Gastric dysrhythmias have been suggested to be indicative of abnormal gastric motility, but a clear cut diagnostic relationship to any disorder has not been established. In addition, a reliable connection between gastric emptying, motility, and electrogastrography has not been demonstrated. EGG is not able to detect any specific disease, and only provides insight when used in conjunction with other tests [114].

1.3.7 Wireless Motility Capsule The wireless motility capsule is a swallowable pill that detects pH, pressure, and temperature as it travels throughout the gastrointestinal tract [115]. The pill transmits recorded information to a wireless receiver that the patient must wear during the test, that stores the data along with a timestamp of when it was received. The pill is designed to travel and record information through the entire GI tract, and time spent in each distinct region can be determined by correlating the data to physiological parameters. In the case of the stomach a sharp increase in pH and pressure signifies a transition to the duodenum of the small intestine [116, 117]. The capsule is typically too large to pass through the pyloric sphincter during digestion, and instead leaves the stomach in phase III of the interdigestive MMC [118]. The benefits of the wireless motility capsule are its portability, lack of radiation, and the absence of a need for expensive imaging equipment. As a result, the Smart Pill could be used in rural settings where access to scintigraphy is limited. However the Smart Pill was not optimized to observe the stomach, and it only provides insight into one instance of gastric emptying that typically occurs well after any food has been expelled from the stomach [118]. In patients with suspected severe gastroparesis it is possible for the pill to never leave the stomach. Studies have shown that taking the pill before or after a meal caused more variation in gastric emptying times than suspected gastric motility disorders [119]. Gastric emptying tests are difficult to

correlate to abnormality as lack of standardization hinders their applicability as a diagnostic tool.

1.3.8 Magnetic Resonance Imaging (MRI) MRI of the stomach can detect wall motion and intraluminal volume change, and as such has been proposed as a radiation free alternative to scintigraphy [120]. Further clinical validation of this technique is required, as MRI is predominantly used for research purposes [121]. To date MRI has not been validated to the same extent as scintigraphy or stable isotope breath tests. The procedure involves the patient ingesting a meal with an MRI marker, and being imaged every 15 minutes. This limits the patient to the supine position, disrupting gravitational effects on digestion. In addition, patients are also required to hold their breath during imaging to minimize motion artifacts [121]. The most dramatic drawback of this technique is the cost and limited availability of the equipment required. A summary of commercially available gastric motility and emptying monitoring technologies is presented in Table 1-2.

Table 1-2: Summary of available technologies to assess gastric motility and emptying

and and

in in

low

and and

MRI

agent

ailability, ailability,

Gastric Gastric

Imaging

contrast contrast

cost and and cost

machine machine

research research

Excellent primarily

Magnetic Magnetic

emptying Very high

Resonance Resonance

volume volume

applications

in in

antral antral

Gastric Gastric

motility motility motility

Unclear Unclear

Capsule

Motility Motility

receiver receiver

Wireless Wireless Wireless

pressure

for gastric forgastric

significance significance

capsule and and capsule

applicability

periods limit limit periods

One time use use time One

interdigestive interdigestive

emptying and and emptying

can can

No No

EGG EGG

states

device device

Gastric Gastric

activity

disease disease

motility

Electro Unclear

electrical electrical

to gastric togastric

recording recording

for gastric forgastric

not detect not detect

stationary)

significance significance significance

(ambulatory (ambulatory

motility,

gastrography

and and and

test

Gastric Gastric

Limited

Balloon

ation of ation

balloon, balloon,

Barostat Barostat Barostat

pressure

tone and and tone

perfusion perfusion

recording recording

caused by by caused

discomfort discomfort

equipment equipment

accommod

Intubation, Intubation,

id id

and and

State State

Good Good

skilled skilled skilled

orsol

requires requires

catheter catheter

pressure

Perfusion Perfusion

technician technician

intubation

technician, technician,

Technically Technically

Antral tone tone Antral

equipment, equipment,

Manometry

challenging, challenging,

and support support and

ray ray

re to to re

Gastric Gastric limited

Contrast Contrast

recorder

to gastric togastric

emptying emptying

radiation, radiation,

Limited in in Limited

emitter and and emitter

Exposu

applicability

agent, X agent, applicability

Fluoroscopy

Test

tone

meal

states, states,

Breath Breath Breath

Gastric Gastric certain

disease disease

isotope isotope

examine examine

vials and and vials

does not does not

Excellent Excellent

emptying emptying

collection collection

Limited in in Limited

for gastric forgastric

tone

gastric gastric

d meal d

Gastric Gastric

Gamma Gamma

camera, camera,

examine examine

does not does not

standard

emptying

radiation, radiation,

emptying, emptying,

radiolabelle

Exposure to to Exposure

current gold gold current

Scintigraphy

Excellent for for Excellent

Function Function

measured

Method of of Method

Equipment Equipment Drawbacks

assessment

requirement Performance

1.4 Aim of this Thesis Existing medical technologies used to monitor gastric motility are severely limited in their ability to predict presence and severity of symptoms in the assessment of two disorders; functional dyspepsia and gastroparesis. In this Thesis the newly proposed, minimally invasive pill-based Transcutaneous Intraluminal Impedance Measurements (TIIM) are presented conceptually and in acute animal studies. The aim of this Thesis is to:

o Design and test a prototype of a pill-based TIIM transducer for use in laboratory and acute animal experiments o Validate pill-based TIIM measurements in a comparative assessment with invasively implanted force transducers physically attached to the serosa of the stomach in acute canine experiments o Design and test a prototype of a TIIM receiver and data logging system in laboratory experiments in order to facilitate future ambulatory testing

Chapter 2: Transcutaneous Intraluminal Impedance Measurements (TIIM) as Enhanced Electrogastrography 2.1 Concept The idea behind using electricity to understand the body is not new, and has been used experimentally since the 1600’s and clinically since the 1800’s. Initially this understanding was limited to the heart, which in itself is a pump, and provides a clean synchronous electrical signal that correlates with contractions. As human electrophysiology was further understood, inherent electrical activity was measured of every major nerve and muscle group, including the stomach. While the stomach shares many of the functions and properties of the heart, the stomach is an asynchronous pump. There is inherent electrical activity within the stomach, measured cutaneously using an electrogastrogram (EGG), however the presence of omnipresent slow waves is not indicative of irregular mechanical contractions [122].

Electrical Impedance Tomography (EIT) has been proposed as a non-invasive technique to visualize the interior of the body [123]. The electrical conductivity or resistivity of the body is reconstructed from the surface potential developed by an injected current measured by an array of cutaneous electrodes. The path the current takes is dependent on the homogeneity of the medium in question, and regions of inhomogeneity will affect the skin potentials and can be reconstructed using complex algorithms. Unfortunately, EIT is an ill-posed nonlinear inverse problem [124], and these drawbacks have limited its clinical applicability [125]. While EIT offers limited practicality the underlying concept of investigating the impedance of the body offers other interesting insights.

The TIIM technique is an extension of EIT in that it involves measuring the resultant skin potential evoked by an oscillating current source, but instead of reconstructing a two dimensional planar image TIIM only measures the relative attenuation dynamics of the signal detected at the skin. The TIIM signal originates from within the stomach, and passes through several non-homogenous layers of tissue to induce a potential voltage at the skin. The signal

will still travel through the inhomogeneous tissue that hampered the outcomes of electrical impedance tomography, but since this tissue is largely static relative to the dynamic muscle contractions of the stomach it has little to no effect on TIIM, which is presented as normalized around a baseline to emphasize impedance fluctuations.

In order to measure impedance there are two key components that need to be considered; the combined resistive and capacitive nature of tissues within the body. Both are dependent on the volume and shape of interrogated tissue, which can be considered for simplification purposes as a dielectric. Since a DC signal cannot detect dynamic capacitance an oscillating square wave signal was used, as in the case of electrical impedance tomography. Previous literature demonstrated that 50 kHz was an ideal frequency to minimize interference with regular muscle, nerve, and organ function, while enabling effective transmission through the body [126, 127]. Body conduction of small low frequency signals has been clinically proven to be safe, and has found application in other GI related technologies. The South Korean company Intromedic has developed and implemented digital communication using body conduction to transmit images from capsule endoscopes at approximately 300 kHz [128]. In the case of TIIM, instead of detecting a digital signal the analog tissue attenuated dynamics are measured from a constant AC power source. A simplified electrical model of the body is provided in Figure 2-1, in which major tissue groups are represented as single resistors. In this model ZTissue and ZSkin remain constant, while ZMuscle is dynamic. In this case changes in voltage across ZSkin are a result of the changing ZMuscle. In the model there are two sources, VTIIM and

VEGG. VTIIM represents the TIIM oscillator, while VEGG represents the slow and spike electrical activity in the stomach. Without the AC carrier signal changes to ZMuscle are not detectable by measuring the voltage evoked by VEGG across ZSkin. This diagram also demonstrates that the reference of the oscillating signal and measurement equipment could only be connected using a catheter. This floating signal facilitates the need for bipolar measurement across ZSkin.

Figure 2-1: A simplified electrical model of the stomach with a TIIM capsule, VTIIM, injecting current into the tissue. ZTissue and ZSkin are constant when compared to ZMuscle, and changes to the impedance of ZMuscle will affect the voltage measured at ZSkin. Without the TIIM capsule, the inherent electrical activity denoted by VEGG is not strong enough to create detectable changes in the voltage across ZSkin. The impedances in this figure denote complex resistive and capacitive components. ZGastric represents the impedance of the surrounding gastric environment.

2.2 Catheter Based Transcutaneous Intraluminal Impedance Measurements; a Feasibility Study A feasibility study was performed of the TIIM technology using a catheter-based system to introduce an electrical signal into the body from within the lumen of the stomach [129]. A 50 kHz 200 mV peak-to-peak sinusoidal signal was generated using a custom external oscillator, and conducted the length of a ZPN-BG-50 Comfortec Z/pH swallowable catheter (Sandhill Scientific, Highlands Ranch, CO) to the electrodes closest to the tip. The choice in catheter was made based on the proximity of the two distal electrodes, ~1.5 cm, that would be closest to simulating a point source of current. The catheter design was tested in two canine models, comparing the effectiveness of TIIM to two force transducers implanted onto the serosa of the

stomach, the signals of which were amplified and conditioned using a custom bridge amplifier. Three pediatric electrocardiographic (ECG) electrodes (Conmed, Utica, N.Y.) were positioned on the skin in a configuration based upon standard EGG measurement in order to record differential surface voltages. The TIIM signals were recorded using a custom-designed multichannel electrogastrograph (EGG, James Long Company, Caroga Lake, NY), which was set to low pass the signals at 0.1 Hz. The signals were digitized using a PCMCIA card DAQ Card-AL- 16XE-50 (National Instruments, Austin, TX, USA) and were monitored, acquired, and analyzed using custom signal processing and visualization software (GAS-6.2, Biomedical Instrumentation Laboratory, University of Calgary, Calgary, Alberta, Canada). Figure 2-2 depicts the setup of this experiment. 30-minute baseline recordings were taken prior to 30-minute recordings of contractions of pharmaceutically invoked contractions using intravenously administered Neostigmine (0.04 mg/kg, APP Pharmaceuticals, Schaumburg, IL).

Figure 2-2: Overall catheter-based experimental setup. 1. Stomach; 2.1. Distal force transducer; 2.2. Proximal force transducer; 3. Connecting wire from force transducers; 4. Transoral catheter; 5. Custom bridge amplifier; 6. ECG electrodes; 7. Receiving wires from ECG electrodes; 8. External TIIM oscillator with isolated ground; 9. Isolated bioelectric amplifier; 10. Analog-to- digital converter; 11. Real time data acquisition software.

The results were evaluated in one-minute motility intervals [130], which effectively demonstrate integrated power of a signal over the course of one-minute. Results from TIIM recordings were compared to the closest corresponding force transducer measurements by calculating Pearson correlation coefficients of the normalized signals. A Pearson correlation coefficient of p<0.05 considered to be significant [131].

2.3 Gastric Retentive Pill-Based Design Following the feasibility study, a pill-based oscillator design was proposed and a prototype was implemented for testing in an eight dog Sham comparison animal study.

2.3.1 Oscillator Design 2.3.1.1 Power Supply Since the capsule was intended for use within the body the choice of power supply had to take safety, power, and space considerations into account. A single silver oxide cell battery (Renata 337, 1.55 V, 8 mAh) provided the energy necessary to drive the oscillator that would interrogate the tissue. This battery is typically employed as a watch battery for its small dimensions and reliability, and silver oxide technology has an excellent safety record in biomedical applications [132], making it an ideal choice.

Lithium polymer batteries offer higher energy density, and have the potential to be incorporated into future iterations of the TIIM capsule [133]. Lithium polymer batteries can be constructed to fit into a smaller sized pill, and can offer longer operating times than silver-oxide batteries, however at the time of the study silver oxide cells are still considered safer than lithium polymer technology [134]. Since the potentially harmful chemicals of the battery are being ingested the toxicity of the battery was the primary design consideration. Proper component choice and efficient electrical design can help extend the life of the battery.

2.3.1.2 Oscillator Once the battery and operating frequency had been selected the next step was to choose an oscillator. Size and power consumption were taken into account when deciding to implement the TS3001 (Silicon Labs, Austin, TX, USA), a programmable oscillator integrated circuit that offered a frequency range from 9 to 300 kHz and uses a maximum of 5.4µA of supply current to maintain operation. The operating voltage of the TS3001 was rated to be between 0.9 and 1.8 V, matching the available voltage from the battery even as it developed internal resistance. The TS3001 has an ultra small footprint as well, only 2 mm x 2 mm, making it 66% smaller than other commercially available CMOS oscillators. The internal circuitry is equivalent to that of a voltage dependant oscillator, where the frequency of oscillation is determined by a voltage divider in which one resistor is controlled by the designer. Silicon labs provided a formula in their data sheet for the chip to calculate the resistor required to control the output frequency. Rset was chosen to be 2.2 MΩ using equation (2.1), with a desired output frequency of 50 kHz.

106 (2.1) 푭 = 3 9.0.9 ∗ 10 ∗ 푹풔풆풕

Where,

F = the output frequency, kHz;

Rset = the user-set resistor, MΩ;

The frequency of oscillation was confirmed to be 50 kHz using a Tektronix TDS 1002 oscilloscope (Tektronix, Beaverton, OR, USA). The TS3001 features two output pins, one with a fixed duty cycle at 50% on pin 1, and a variable duty cycle that could range between 12 and 90% on pin 3. A 50% duty cycle was chosen for simplicity and to minimize extraneous resistors to set the duty cycle of the programmable pin. One decoupling capacitor was also integrated at the

recommendation of the manufacturer. Using an integrated circuit instead of discrete components decreased the footprint, and reduced the complexity of construction. By utilizing fewer parts there were also fewer locations for potential soldering error or manufacturing defects. The complete design including the battery, oscillator chip, resistor and capacitor fit onto a 10 mm by 6 mm two sided printed circuit board that was custom manufactured.

2.3.1.3 TIIM Capsule Body In order to protect the circuitry, the custom electronics were held in a custom capsule that was made for the prototype models. The capsule consisted of a cylindrical body with two cap electrodes on either end, with an outer dimension of 10 mm in diameter and 19 mm in length, (Figure 2-3). The hollow cylindrical body was made of a machined biocompatible chemically resistant medical grade polyetherimide (PEI) resin (Ultem 2300; Ritter GmbH, Schwabmunchen, Germany). This material was selected for its corrosion resistance, dielectric insulation, and high strength to weight ratio [135]. The inner diameter of the cylindrical body was 8.36 mm, allowing the custom electronics to fit. On either end of the body the last 2.5 mm had recessed threading to allow custom made caps to be tightly screwed into place. The main constraint on the size of the capsule was the presence of embedded electronics, as well as the need for gastric retentive polymers surrounding the capsule but with in the dissolvable pill body. The custom caps were made of machined copper that was threaded to closely fit the body in order to be able to seal the contents of the pill. The inside surface of the copper caps was soldered to thin wire connected to output vias on the circuit board before being screwed together. In order to ensure that body fluids did not enter the capsule the two caps were screwed onto the shell with a layer of ultra-thin polytetrafluoroethylene (PTFE) film in order to ensure a high quality waterproof seal between threads. Figure 2-3 shows the full design of the internal circuitry and body of the TIIM capsule.

Figure 2-3: The oscillator circuit, printed circuit board layout, and TIIM capsule body. Dimensions of the circuit board were 10mm x 6mm, which allowed it to fit inside the capsule body that had dimensions 19mm x 10mm. Size of the board was constrained by component choice, as well as the need for swelling polymers surrounding the capsule body.

2.3.1.4 Power Consumption After construction short circuit current analysis was performed by connecting a copper wire [22 AWG, Alpha Wire, Elizabeth, NJ, USA] between the two electrodes and measuring the current running through the wire using a hall effect sensor [CC-65, Hantek, Qingdao City, Shandong Province, China]. The maximal output current was found to be 0.3 mA peak to peak, and the frequency of the oscillation was further confirmed to be 50 kHz. With this in mind the worst case in-vivo average power consumption was found to be 0.328 mW using equation (2.2).

푷푨풗품 = 푰풓풎풔 ∗ 푽풓풎풔 (2.2)

Where,

PAvg = the average power consumed by the oscillating signal, mW;

Irms = the root mean square of the output current, mA;

Vrms = the root mean square of the output voltage, V;

The battery has a capacity of 8 mAh when supplying 15.6 µA, however the designed circuit draws 0.3 mA. The maximum operating time of the capsule was estimated to be 26.6 hours using equation (2.3), knowing that it can be expected to operate with a lower capacity.

푪 푻 = (2.3) 푴풂풙 푰

Where,

TMax = the maximum operating time, hours;

C = the battery capacity, mAh;

I = the output current, mA;

The IEEE has made recommendations for the maximum time-varying current for signals varying between 1 Hz and 100 kHz for the general public that can be applied to the TIIM signal [136]. This recommendation is made on the basis of minimizing interactions with nerve and muscle tissue, and is significantly higher than when considering direct current sources as the frequency is much higher than that of depolarization of the tissues. Within their recommendation they provided a formula for the maximum time-varying contact current. At 50 kHz, the maximum allowable current was found to be 10 mA using equation (2.4).

푰 = 0.2 ∗ 풇 (2.4)

Where,

I = the maximum allowable current, mA; f = the frequency used, kHz;

Comparing the maximum allowable current to the maximum output current of the TIIM oscillator it is 33 times lower, making this pill well within the reasonable range of current exposure for the general public, even when considering the substantial DC component of the signal. Higher currents could be used in future TIIM designs, however, a power source with much higher power density and discharge rates would be required to be able to meet these higher requirements.

2.3.1.5 Gastric Retention In order to produce long term results and prevent the expulsion of the oscillator pill into the duodenum, a gastric retentive enclosure was incorporated into the prototype models. The oscillating TIIM capsule was embedded in 0.15g of dry, non-toxic, hydrophilic, crosslinked polyacrylate polymer granules contained inside a nonwoven, high permeability 20-gsm polyvinyl alcohol (PVA) mesh. The polymer granules are able to immediately absorb and retain hundreds of times their weight in water, and swell to 30-50 times their dry size in gastric liquid. Once the polymers have swelled they are unable to dissolve due to the three dimensional crosslinked structure [137], making them an ideal choice for gastric retention when held in place by a mesh structure. The PVA mesh was designed to ensure that upon expansion the whole structure would exceed 1.5 cm in any direction, but not be larger than 4 cm in any direction, with a fully expanded volume of less than 30 ml, which is well below the threshold of perception [138]. When pressure is applied the polymer granules linearly retain less water, allowing a certain degree of compliance of the structure to the mechanical activity of the stomach. The mesh was also chosen due to its permeability to fluids, which enable the polymer granules to make adequate contact with the gastric juices. The PVA mesh is biodegradable, and will disintegrate within 2-3 days in order to avoid obstruction, and similar gastric retentive technologies have been demonstrated to withstand digestive antral pressures, and produce no adverse mucosal impact or evacuation obstruction issues [139]. The PVA mesh can be immediately disintegrated on demand via the consumption of hot (>50o C) water.

2.3.1.6 Encapsulation In order to orally deliver the TIIM capsule and gastric retentive enclosure to the stomach the prototype models were encased in a hard gelatin shell capsule (Torpac Inc., Fairfield, NJ, USA). The gelatin shell was selected because it dissolves very rapidly (2-4 minutes) in water, thus facilitating faster expansion of the expanding polymers. A size #13 capsule was chosen taking into account the animal testing requirements of this study. The size of the gelatin capsule directly affected how much expanding polymer could be integrated into the design, as the size of the TIIM capsule was determined from the internal electronics. For the present study the emphasis was on demonstrating the effectiveness of the electronics. Figure 2-4 presents an exploded image of the pill, as well as the order of assembly.

Figure 2-4: Breakdown of the TIIM gastric retentive pill. 1. Oscillator circuit; 2. Capsule body; 3. Assembled capsule; 4. Superabsorbent granules; 5. Capsule and granules inside a liquid- permeable mesh; 6. Dissolvable pill containing the mesh enclosed capsule; 7. TIIM gastric retentive pill; 8. Pill expanded in water; 9. Test dish. The oscillator circuit was first sealed into the assembled capsule, and added to superabsorbent granules held in a liquid-permeable mesh. This was inserted into a dissolvable pill, which completely disintegrated in the stomach after 10 minutes, allowing the granules to swell. A mm scale is provided on the left of the image.

2.4 Signal Acquisition, Conditioning, Digitization, Processing, and Data Logging 2.4.1 Signal Acquisition The skin potential evoked from the induced TIIM signal was measured via repositionable Ag/AgCl ECG electrodes (Cleartrode, Conmed, Utica, N.Y., USA). A single TIIM channel is recorded as the potential difference between two electrodes measured relative to the voltage of a ground electrode placed away from the TIIM transducer. With signal processing this bipolar signal can demonstrate the power of the signal, which is hypothesized to correlate highly with contractile activity. Electrodes should be placed in a similar configuration to that used in EGG [111] in order to pick up an adequate amount of signal in order for meaningful information to be recorded.

2.4.2 Signal Conditioning The raw analog signals from the ECG electrodes were amplified and filtered using a custom-designed multichannel electrogastrograph (EGG, James Long Company, Caroga Lake, N.Y. USA), which measured the voltages relative to the ground electrode. The custom electrogastrograph is a battery powered device (12V, 2.3 Ah, PS-1223, Power-sonic Corporation, San Diego, CA, USA) that can process 16 different channels. Each channel has gain options of 1000X, 10 000X, 50 000X, and 100 000X, however the majority of the gain options were redundant as TIIM measures the resultant amplitude of the oscillator instead of the inherent electrical signals. The device also features high and low pass filters optimized for recording gastric electrical activity. The high pass filtering options are DC, 0.015 Hz, 0.03 Hz, and 0.3 Hz while the low pass filtering options are 0.1 Hz, 0.5 Hz, 100 Hz, and 200 Hz. These filtering options can reduce unwanted physiological noise, such as breathing or cardiac rhythms as these processes occur at much higher frequencies than gastric contractile activity.

2.4.3 Signal Digitization After being amplified and filtered the TIIM signal was digitized using a PCMCIA DAQ Card-AL-16XE-50 (National Instruments, Austin, TX, USA). This analog to digital converter can record 16 bit voltages between 0 and 5 volts at a maximum rate of 200 kS/s. Since the frequency of gastric contractions was estimated not to exceed 0.2 Hz in canines a sampling frequency of 10 Hz was chosen for simplicity and to avoid aliasing effects. The resolution of this analog-to-digital converter was found to be 0.076 mV ± 0.038 mV using equation (2.5).

푨푹 푨푹 푨 = ± (2.5) 푹풆풔 2풏 2풏+1

Where,

ARes = The analog resolution of the system AR = the full analog range of the ADC n = the number of bits in the ADC

2.4.4 Signal Processing The signals were viewed using custom signal processing and visualization software (GAS- 6.2, Biomedical Instrumentation Laboratory, University of Calgary, Calgary, Alberta, Canada). This software was designed with electrogastrography in mind, and can visualize up to 16 channels simultaneously in real time. Signals are monitored and stored for further analysis.

After signals were acquired they were further processed using a custom program developed using Matlab (MathWorks, Natick, MA, USA) that normalized each signal before calculating one-minute motility indices, and computing Pearson correlation coefficients between the input signals. Normalization is done because of the inhomogeneity of the tissues between the TIIM transducer capsule and whichever electrode recorded the signal. As a result,

TIIM signals are presented in terms of relative units, where the highest value is represented by a 1 and the lowest by a 0. Normalization was performed according to equation (2.6).

푫푶 − 푫푴풊풏 푫푵풐풓풎 = (2.6) 푫푴풂풙 − 푫푴풊풏

Where,

DNorm = the new normalized data point, dimensionless;

DO = the original point to be normalized, dimensionless;

DMin = the minimum data point in the entire set, dimensionless;

DMax = the maximum data point in the entire set, dimensionless;

One-minute motility indices [130] effectively low pass filter the signals, reducing the effect of erroneous outlying data and presenting the general trend of gastric motility. Motility indices are a convenient way to interpret the impedance dynamics of the tissue because they represent the power of the signal [140], which is independent of pill polarity. This is important because the pill is effectively floating within the stomach, and its orientation cannot be predicted so by measuring the power of the signal we can negate this effect. One-minute motility indices are further advantageous because signals from force transducers and TIIM may be out of phase due to the spatial difference in ECG electrodes. In this case if could appear that little to no correlation in activity was present. Motility indices are calculated by taking the sum of squared data points over the course of a minute. In one minute there were 600 samples, as the sampling frequency was 10 Hz. After motility indices were calculated using equation (2.7) they were again normalized according to equation (2.6). This normalization is indicative of the average power, and can be compared across samples even with varying sample rates.

600∗(푖+1) ퟐ 푴풊 = ∑ 푫풋 (2.7) 푗=600∗푖+1

Where,

Mi = the motility index M calculated for minute i, dimensionless;

Dj = the normalized data at point j, dimensionless;

Once the general trends of gastric motility are established by the motility indices, Pearson correlation coefficients can be computed in order to compare the different signals, and provide a metric for validation. Pearson correlation coefficients were chosen for statistical validation because they indicate a measure of linearity between two signals. Pearson correlation coefficients with p<0.05 were considered to be significant [131].

Further processing of the signals was done to assess the contractions per minute visible in each respective channel. Dominant peaks of the frequency for each measuring modality (basal and post-neostigmine) were subjected to a comprehensive statistical analysis using the paired Student’s t-test [140] to evaluate the relationship between the frequency dynamics of TIIM, sham and force transducer measurements. Frequency analysis was done using the Fast Fourier Transform, implemented in the custom visualization and analysis GAS-6.2 software used previously.

2.5 Development of a portable TIIM receiver In order to facilitate future testing of the TIIM technique a portable TIIM receiver was developed based on the analog conditioning options available in the bioelectric amplifier that were most suitable for measuring the TIIM signals in canine studies, in which an amplifier gain over 1000 caused clipping during periods of higher motility. The requirements of the receiver

were to have an array of inputs to accommodate two TIIM channels, multiple analog amplification and filtering options that could be accessed on the device, and a microcontroller and Bluetooth transceiver to facilitate the transmission of data to any Bluetooth enabled device such as a smart phone or personal computer on which visualization and storage can take place.

2.5.1 Inputs The portable TIIM receiver was designed to have two bipolar inputs, but this could become two unipolar inputs by connecting the negative input to ground. The inputs are standard female micro-banana plug receptacles (Johnson 105-0753, Cinch Connectivity Solutions, Waseca, MN, USA) that match the male end of the ECG electrode clips that were used. These are color coded with red being positive, black being negative, and green being ground to facilitate ease of use.

2.5.2 Amplification and filtering The incoming signals are first met by an instrumentation amplifier (INA333, Texas Instruments, Dallas, TX, USA) which is applicable to both bipolar and unipolar signals depending on the application. The gain of this stage was fixed to be 1.05 by determining RG to be 200 MΩ using equation (2.8) to reduce noise throughput.

100푘Ω 푮 = 1 + ( ) (ퟐ. ퟖ) 푹푮

Where,

G = the gain of the instrumentation amplifier, dimensionless;

RG = the feedback resistor between the two inputs of the instrumentation amplifier, Ω;

The signal then passes through a variable gain and filtering stage. Using a 1st order low pass active filter design in which the input resistance and feedback capacitance can be changed enables a user to adjust the acquisition to parameters that they are interested in. The active filter was implemented using a low power CMOS operational amplifier (OPA333, Texas Instruments, Dallas, TX, USA), which was selected for its current consumption of 17 µA, and the low noise properties of CMOS technologies. The low pass filtering options draw upon those of the custom made electrogastrograph that was utilized during feasibility testing. Options can be selected between using surface mount DIP switches (8 position DIP switch, TE Connectivity, Schaffhausen, Switzerland). By using the DIP switches up to 16 possible gains and filtering constants can be attained, by combining which switches are on or off. The analog amplification and conditioning stage of the TIIM receiver is shown in Figure 2-5.

Figure 2-5: The analog amplification and conditioning circuit implemented in the portable TIIM receiver. This circuit is implemented twice to provide two TIIM channels on the prototype receiver. Variable gain and filtering parameters can be set using the DIP switch between the two stages. The offset of the output voltage can be adjusted using the 20kΩ potentiometer. In this configuration the gain is set to 1000X with no filtering option selected.

The variability of the gain comes from the input resistor of the final stage. By applying different combinations of switches gains ranging from 1X to 1111X can be achieved. A reference chart for the gain and filter switch settings is presented in Figure 2-6. By turning off all resistor switches the device enters a calibration mode, in which the signal does not transmit to the final stage, and the offset can be set to zero before recording if so desired. The gain of the final stage can be calculated using formula (2.9).

푹푭 푮 = ( ) (ퟐ. ퟗ) 푹푰

Where, G = the gain, dimensionless;

RF = the feedback resistor, Ω;

RI = the variable input resistor, Ω; In order to control the cut off frequency of this stage without compromising the gain, a variable capacitor array produces filtering constants from 0.087 Hz to 96.7 Hz. Capacitors were selected based on availability in 0805 surface mount packages, with the aim of providing filtering constants similar to the ones found on the custom electrogastrograph. Filtering switch configurations are provided in Figure 2-6. The cutoff frequency of the low pass filter can be calculated using formula (2.10).

1 푭풄 = (ퟐ. ퟏퟎ) 2 ∗ 휋 ∗ 푹푭 ∗ 푪푭

Where,

FC = the cutoff frequency, Hz;

RF = the feedback resistance, Ω;

CF = the variable feedback capacitance, F;

Figure 2-6: DIP Switch Reference Chart

The total gain of the system can be calculated by multiplying the gain from the instrumentation amplifier, 1.05, with the user selected gain, 1- 1111. The intended use of the TIIM receiver was to have the DIP switches set before any recording. Each channel also includes a voltage offset on the filtering and amplification stage of the circuit. This is facilitated via a 20 kΩ potentiometer (Panasonic, Kadoma, Japan) with the wiper on the positive input of the CMOS op-amp and either end tied to the power supply. This surface mount potentiometer can be adjusted using a small screwdriver, and is also intended to be left alone during recording after being set appropriately.

2.5.3 Microcontroller and Bluetooth An RFduino 22102 chip (RF Digital Corp USA, Hermosa Beach, CA, USA) was implemented as an integrated Bluetooth radio with built in analog to digital conversion. The chip operates at 16 MHz, and features 8 kb of ram and 128 kb of flash memory on which to store instructions. The analog to digital converter is 10 bit, and can process voltages between 0 and 3 V. The resolution of this system was found to be 3mV ± 1.5 mV using equation (2.5). This resolution was lower than the resolution of the ADC card used during the canine tests, as the chip was chosen for its ease of use and integrated antenna more so than the resolution of its ADC.

2.5.4 Battery and portability considerations Low power components were selected with portability in mind. A CR2032 battery is enough to power the entire circuit (225 mAh, 3V, Panasonic, Kadoma, Osaka Prefecture, Japan) for an extended period of recording and transmission of data at a rate of 10 Hz. At this data transmission frequency, the circuit will draw under 9 mA of current, including 0.5 mA for a power indicator LED (5mm Green LED, Rohm Semiconductor, Kyoto, Japan). Further voltage regulation should be implemented in future designs, as well as higher capacity rechargeable batteries to offer longer term operation for users. Using equation (2.3) the worst case battery life will be 25 hours, closely matching the pill-based oscillator. All the components were arranged onto a 7 cm x 6 cm custom printed circuit board, shown in Figure 2-7.

Figure 2-7: The TIIM receiver and its components.

2.6 Laboratory Testing 2.6.1 Laboratory Testing of TIIM capsule Initial testing of the TIIM capsule was done using tripe from the reticular stomach, which served as a suitable tissue analog for preliminary testing. A prototype of the capsule was inserted into a piece of tripe which was then folded over twice to simulate multiple layers of tissue. Saline (0.9 % w/v of NaCl) was liberally applied between layers to mimic the composition of the body. Repositionable ECG electrodes were placed over the pill and the voltage relative to a ground electrode placed further away from the capsule using the acquisition method and system described. After the signal was established the tripe was manually palpated and changes in amplitude observed.

2.6.2 Laboratory Testing of Portable TIIM Receiver The portable TIIM receiver was designed and tested after the success of the animal studies. It was designed to amplify and filter acquired signals using similar inputs and parameters as the custom electrogastrograph but in a much smaller form. The Bluetooth capabilities of the TIIM receiver were used to transmit data using the Bluetooth 4.0 protocol to a nearby computer where the incoming signal could be visualized and assessed. In order to recreate the in-vivo TIIM signal the same DAQ card that was used to acquire the signal originally was used as a digital to analog converter that output the voltages to the TIIM receiver. Voltages were delivered to an attenuation circuit at 10 Hz to mimic the original test conditions.

In order to simulate the same test conditions, the signal was first attenuated using two resistors to make a voltage divider. Resistor values were chosen to attenuate the voltage by a factor of approximately 1000 using formula (2.11) [141].

푹ퟏ 푽풐풖풕 = 푽풊풏 ∗ (2.11) 푹ퟏ + 푹ퟐ

Where,

Vout = the output voltage of the voltage divider, V;

Vin = the input voltage of the voltage divider, V;

R1 = the resistor across which Vout is produced, Ω;

R2 = the resistor to limit the current required to produce Vout, Ω;

The portable TIIM receiver was set to amplify the signal by close to the amount attenuated, 1050X, however the low pass filter was disabled to minimize distortion of the incoming signal. The signal produced by the digital-to-analog card was a recreation of an already filtered signal, so any further filtering would introduce error. Following signal acquisition and storage the incoming signal was compared to the original signal using the same Matlab code that calculated Pearson correlation coefficients between two signals in order to

assess whether the signal from the TIIM receiver accurately represented the signal produced by the digital to analog converter. The overall experimental setup is shown in Figure 2-8.

Figure 2-8: The experimental setup to assess the portable TIIM receiver, as well as the circuit diagram of the attenuation circuit.

2.7 Animal Testing Experiments were performed on eight mongrel dogs under CCAC protocol SHC11R-03, six of which were female, that had a mean mass of 23.8 kg ± 3.3 kg. Four were administered an active TIIM capsule, while the rest were given a deactivated (battery removed) capsule. The dogs were vaccinated and dewormed as per the Canadian Veterinary Medical Association’s recommended yearly protocol regime. Vaccines included Canine Distemper/adenovirus, Type 2

Parvovirus/Bordetella/Rabies. Drontal Plus (Bayer HealthCare LLC, Shawnee Mission, KS, USA) was used as an oral dewormer. Each animal underwent a physical examination by a board- certified veterinarian and was found to be in good condition. After 24 hours of fasting and 12 hours of water deprivation, each animal transorally ingested a single capsule as described above with 500 cc of room temperature water (21.0o C). The pill swelled to its maximum size in the stomach within 15 minutes after ingestion to dimensions exceeding 1.5 cm in any direction, and subsequently was unable to pass the pyloric sphincter even when subjected to pharmacologically induced propulsive peristalsis.

Each animal underwent an induction with an intravenous injection of thiopental (Thiotal 15 mg/kg IV, Vetoquinol Canada, Lavaltrie, QC, Canada) and was subsequently maintained on inhalant isoflurane and oxygen (Halocarbon Laboratories, River Edge, NJ, USA) with a vaporizer setting of 1-3%. The anesthesia was chosen because it did not influence gastric neurotransmitters, and as such would not affect gastric contractions [142]. Individually, the animals were then positioned supine, their abdomens shaved, cleaned, and sterilized with alcohol before performing laparotomy via a median incision vertically along the linea alba to gain access to the stomach.

After the incision the location of the ingested pill in the stomach was verified endoscopically using an EPK-700 veterinary endoscope (Pentax, Tokyo, Japan), and the voltage developed on the serosa of the stomach was measured using an oscilloscope (Tektronix, Beaverton, OR, USA) to confirm the presence of an activated or deactivated pill. After this verification, two 90W24 force transducers (RB Products, Stillwater, MN, USA), specifically designed for gastric motility monitoring were surgically sutured to the serosal side of the antral stomach along the gastric axis [143]. The first force transducer was positioned 1-2 cm from the pylorus, and the second was affixed proximally 5-6 cm from the pylorus, along the gastric axis. The mesenteric innervation and the blood supply of the stomach were carefully preserved. The internal position of the TIIM oscillator and the serosal position of the force transducers is shown in Figure 2-9.

Figure 2-9: The serosal view of the stomach showing the placement of the force transducers (A), and the internal view of the stomach showing the position of the TIIM capsule and gastric retentive mesh.

The signals from the force transducers were amplified using a custom-designed multichannel bridge amplifier, and digitized using a PCMCIA DAQ Card-AL-16XE-50 (National Instruments, Austin, TX, USA). The force transducer signals were monitored and analyzed with custom-designed signal processing and visualization software (GAS-6.2, Biomedical Instrumentation Laboratory, University of Calgary, Calgary, Alberta, Canada). Once the force transducers were in place, their functionality was verified mechanically by manual palpation of variable strength, and the offsets and gains were calibrated accordingly for maximal sensitivity. The intragastric position of the pill was then verified mechanically by palpating it to ensure that it had not been compromised during surgery.

Following the FT implantations, the abdomen was closed, and after appropriate skin cleaning and preparation, three pediatric ECG electrodes (Conmed, Utica, N.Y., USA) were placed cutaneously over the stomach along the abdominal projection of the gastric axis, with a ground electrode positioned close to the left hip of the animal [144]. The position of the electrodes was similar to the one associated with impedance epigastrography, since previous studies have suggested optimal electrode placement [145].

The TIIM signals were measured according to the method described. The amplification of the custom electrogastrograph was set to 1000X, in order to utilize the range of the analog to digital converter. The cut-off frequencies of the bandpass filter of the custom electrogastrograph were set to the commonly used 0.03 - 0.1 Hz following the hypothesis that gastric motility signals in the animals will not exceed 6 cycles-per-minute (cpm) [146] and will amplitude-modulate the intraluminal oscillator frequency of 50 kHz, the latter acting only as their carrier. The 0.1-Hz low pass filter would thus act as a demodulator for this transcutaneous signal transmission, and would prevent higher frequency electrophysiological and mechanical processes (e.g. electrocardiographic activity and respiration) from interfering with the signal originating from within the stomach. The signals were then digitized using the same PCMCIA card DAQ Card-AL-16XE-50 (National Instruments, Austin, TX, USA) simultaneously with the force transducer signals, and were subsequently monitored and stored for further analysis using the same custom software. The overall setup of this experiment is shown in Figure 2-10.

Figure 2-10: Overall gastric retentive pill-based experimental setup. 1. Stomach; 2. Gastric retentive mesh; 3. TIIM capsule; 4. Implanted force transducers; 5. Connecting wire from force transducers; 6. Custom bridge amplifier; 7. Cutaneous electrodes; 8. Receiving wires from ECG electrodes; 9. Isolated bioelectric amplifier; 10. Analog-to-digital converter; 11. Real-time data acquisition software.

Immediately after the experimental setup was completed [144], a baseline recording was performed with no pharmacological stimulant for 30 minutes. Following this recording, bolus neostigmine (0.04 mg kg-1, APP Pharmaceuticals, Schaumburg, IL) was administered intravenously as a smooth muscle stimulant to invoke contractions [143]. Thirty minutes of post-neostigmine recordings were subsequently obtained. The total recorded time from each animal was 60 minutes; 30 minutes in the basal state, and 30 minutes in the post-neostigmine state, with a one-minute time interval between them for the intravenous (IV) administration of the bolus neostigmine. The overall experimental procedure is shown in Figure 2-11.

Figure 2-11: The overall experimental procedure.

At the end of the experiments the animals were sacrificed by an IV injection of Euthanyl, 480 mg/4.5 kg (Bimeda-MTC Animal Health Inc., Cambridge, ON, Canada). Subsequent retrieval of the expanded pill was performed in order to verify its retention within the stomach, and to confirm the presence of the signal in the active TIIM pills or the lack thereof in the inactive sham pills using an oscilloscope. The post-administration volume of each gastric-retentive pill was measured to quantify expansion dimensions. This study was approved by the Veterinary Sciences Animal Care Committee, University of Calgary, Calgary AB, Canada.

Chapter 3: Results 3.1 Laboratory Testing 3.1.1 Laboratory Testing of the TIIM Capsule The activated TIIM capsule was able to demonstrate reliable signal transmission through multiple layers of tripe, and variable attenuation of the signal was observed during manual palpation. During this test the frequency of the square wave emitted by the TIIM capsule was verified to be 50 kHz while the tripe was not being palpated. A sample recording from the oscilloscope is shown in Figure 3-1 (A). Upon palpation attenuation of the signal was observed, and appeared to be affected by mechanical activity of the tripe. Care was taken to ensure that the changes in amplitude were not a result of movement of the pill, but rather were caused by changes in the thickness and density of the compressible tripe. A sample recording from the oscilloscope during palpation is shown in Figure 3-1 (B). In order to assess the effect of palpation the time scale was adjusted to 5 seconds per screen, so as to better match the frequency characteristics of human activity.

Figure 3-1: An oscilloscope reading from cutaneous electrodes attached to tripe to verify the frequency of the TIIM oscillator inside (A). Palpations were visible as amplitude modulations (B).

3.1.2 Laboratory Testing of the Portable TIIM receiver Laboratory testing of the portable TIIM receiver demonstrated that the recordings taken using the prototype device were significantly correlated to the output signal generated using a recording taken with the data acquisition equipment utilized during animal testing. A Pearson correlation coefficient of 0.99 was calculated between the normalized input and output motility indices, meeting the criteria (p<0.01) required to demonstrate significance. The portable receiver had lower ADC resolution, by a difference of six bits. This discretization was only observable upon very close inspection, and had minimal effect on the one-minute motility indices obtained from the TIIM receiver data. The two signals were not completely identical because of possible noise contamination within the testing apparatus, as well as the slight discretization. Figure 3-2 shows both the raw signals as well as their one-minute motility indices for the thirty-minute recording.

Figure 3-2: Combined plot of the pre-attenuated signal, portable receiver recorded signal, and their respective one-minute motility indices. The input signal was taken from a 30-minute recording obtained during animal testing.

3.2 Animal Testing 3.2.1 Gastric Retention Before the implantation of the force transducers an oscilloscope was used to detect the 50 kHz signal, or lack thereof being emitted by the ingested TIIM capsule. Oscilloscope leads were placed in electrical contact with the serosa of the stomach over the pill, the position of which was approximately determined by manual palpation. A sample oscilloscope recording for an active pill is shown in Figure 3-3 (a), while Figure 3-3 (B) shows a recording of an inactive pill. The voltage and time scales of the oscilloscope were set to the same settings used during laboratory tests using tripe.

Figure 3-3: An oscilloscope reading from the gastric serosa prior to the force transducer implantation verified the presence of an activated TIIM pill (A). The sham pills did not demonstrate any signal (B).

The expanded pill was retrieved from each animal after experimentation, in order to verify the pill’s activity using the same oscilloscope used previously, as well as to assess the expansion dimensions and structural integrity of the gastric retentive enclosure during testing. Each experiment revealed that all the TIIM pills remained either active during testing, and did not fail as a result of liquid exposure or battery failure. Conversely, the sham pills remained inactive, as

expected. The average post-retrieval volume of the pills was 12.1 ± 0.4 ml, and had dimensions exceeding 1.5 cm in all directions. The presence of the intact pill within the stomach at the conclusion of the experiments indicated that the pharmacologically induced contractions were unable to propel the expanded gastric retentive enclosure into the small intestine, nor were the pressures within the stomach high enough to rupture the gastric retentive enclosure itself.

3.2.2 Motility Indices, Pearson Correlation Coefficients, and Contractions per Minute The testing was divided into two sections, baseline and post-neostigmine, and results were recorded and labelled accordingly. During baseline tests there was varying evidence of spontaneous contractile activity as measured by the force transducers, as evidenced in visual comparison of Figure 3-4 and Figure 3-5. Following the introduction of neostigmine several gastric motility factors dramatically increased, notably the frequency, regularity, and amplitudes of contractile activity. A typical simultaneous force transducer and TIIM recording for an active capsule is shown in Figure 3-4, along with the post processed one-minute motility indices in parallel. Thirty minutes of baseline activity and thirty minutes of post-neostigmine activity was recorded and normalized together, in order to put each recording into perspective with respect to the maximum and minimum of both states combined. In both states there were statistically significant (p<0.01) correlations demonstrated between the one-minute motility indices of TIIM and force transducer signals. The more distal or proximal cutaneous electrode combinations were compared to their corresponding force transducer. In the case of the battery removed sham pill the results showed no significant correlations between respective motility indices. Another typical simultaneous recording is shown in Figure 3-5, for the case of the sham pill. Table 3-1 summarizes the Pearson Correlation Coefficients of the one-minute gastric motility indices, divided into baseline or post-neostigmine, and active or sham capsule.

Table 3-1: Averaged Pearson correlation coefficients (PCCS) of the one-minute gastric motility indices (GMIs) per state per capsule type.

PCCs proximal PCCs distal FT-proximal FT-distal Modality State p value p value Cutaneous Cutaneous GMIs GMIs Baseline 0.763 ± 0.2 < 0.01 0.674 ± 0.47 < 0.01 TIIM After capsule 0.731 ± 0.12 < 0.01 0.734 ± 0.14 < 0.01 neostigmine Baseline 0.160 ± 0.03 > 0.1 0.071 ± 0.02 > 0.1 Sham After capsule 0.113 ± 0.09 > 0.1 0.051 ± 0.03 > 0.1 neostigmine

Frequency spectral analysis was performed in order to assess the contractions per minute of each signal. As in the case of Pearson correlation coefficients distal and proximal TIIM recordings were compared to the corresponding force transducer. The dominant frequency peaks revealed statistically significant relationships between TIIM and force transducer recordings. The sham study revealed substantial dissociation between the dominant frequency peaks, particularly during the baseline period in which the limited spontaneous gastric activity was sporadic and irregular. Accordingly, there was no significant relationship between the sham study, which can be effectively be considered an electrogastrography study, and the force transducers. The averaged values of the dominant frequency spectra (0.03-0.1 Hz) are presented in Table 3-2. Comparative statistical evaluation of the frequency dynamics of the dominant spectral peaks is presented in Table 3-3.

Table 3-2: Averaged cycles per minute (CPM) of the raw force transducer (FT) and cutaneous recordings per state per capsule type.

Modality State Channel CPM

FT 2.38 ± 1.2 Baseline TIIM 2.42 ± 1.27 TIIM capsule FT 3.55 ± 0.94 After neostigmine TIIM 3.58 ± 0.95

FT 2.65 ± 1.15 Baseline Sham 3.94 ± 1.67 Sham capsule FT 3.84 ± 0.91 After neostigmine Sham 4.12 ± 1.56

Table 3-3: Statistical comparison between the dominant frequency peaks of the FT and TIIM/Sham recordings using paired Students t-test. *(p < 0.05).

Modality State P value Baseline (FT TIIM) 0.048* TIIM capsule After neostigmine (FT TIIM) 0.049* Baseline (FT TIIM) 0.92 Sham capsule After neostigmine (FT TIIM) 0.33

n the baseline (0 state the baseline n

minute motility indices for an active active i pill forindices motility minute an

3600). A thick vertical line denotes the administration of neostigmine. the denotes administration line thick vertical A 3600).

–

neostigmine administration (1800 administration neostigmine -

: Combined plot of the raw signals and the one plot and the : signals raw of Combined 4

- 3

Figure Figure 1800)post and relative in displayed and U.). Measurements normalised are (R. units

active pill in the baseline the in baseline pill active

3600). A thick vertical line denotes the administration of of the verticaladministration denotes line thick A 3600).

–

minute motility indices foran indices motility minute

neostigmine administration (1800 (1800 neostigmine administration

Measurements are normalised and displayed in relative units (R. U.). (R. units displayed relative in and normalisedare Measurements

: Combined plot of the raw signals and the the one and the signals raw plot Combined of :

1800) and post 1800)and

– -

3 tigmine.

Figure state (0 neos

Chapter 4: Discussion The quest to understand the motility of the stomach has been dominated by gastric emptying scintigraphy, but with the rise and subsequent validation of radiation-free C13 breath tests this could soon change. In both cases the information presented by the tests is inherently limited due to the issues surrounding variability of interpretation, as well as standardization and patient positioning[88], and the ties between gastric emptying rates and symptoms generated in patients[48]. With the global rise of diabetes [147, 148] there will be a corresponding increase in abnormal GI function, which further complicates the development and management of the disease [149]. Asymptomatic delayed gastric emptying can be a reality for diabetic patients [47], however, the opposite is also found in cases where symptoms are present without abnormal gastric emptying [150]. In these cases gastric emptying studies are inadequate, as they reveal nothing about the development of symptoms, and may represent a significant cost to patients whose insurance may not reimburse for a second test [88]. Understanding gastric motility is essential to proper management of postprandial glycaemia [151], and the test meal may be a misrepresentation of the actual diet of a diabetic patient. Current gastric emptying studies provide insight into one instance of emptying, making it difficult to observe daily or weekly trends with varying diets. Manometric studies can offer insight into neuropathy or myopathy from the pressures applied from the stomach, but has been criticized for catheter migration and patient discomfort [152]. Trends in medical technologies indicate a shift towards cheaper, more available tests that can provide similar insights.

The wireless motility capsule (WMC) promises comprehensive analysis of the entire GI tract, with the possibility to diagnose multiregional dysmotility [153]. Pressure, pH, and regional transit times are transmitted using radio-waves, and has shown promise in comparative studies with traditional radiography testing modalities [154]. Despite the enthusiasm for the technique, the WMC is only able to demonstrate one instance of gastric emptying during phase III of the migrating motor complex, which is associated with indigestible solids rather than food or liquids

[118]. The WMC, along with capsule endoscopy [155] is representative of a trend towards minimally-invasive office-based testing of gastrointestinal disorders and diseases.

The TIIM technique offers a low-cost minimally-invasive alternative to traditional gastric motility testing, and can be thought of as an enhanced electrogastrogram (EGG). The clinical utility of gastric electrical measurements has so far been limited [156, 157]. The spontaneously generated intrinsic gastric electrical activity, detected and recorded using a traditional EGG, reflects an omnipresent phenomena known as periodic gastric slow waves or electrical control activity [158]. These waves are a result of the cellular Na+ K+ exchange that causes depolarization that is necessary but not sufficient for contractions to occur [159]. For gastric contractions to occur further Ca2+ influx at the cellular level is required, which synchronizes with the electrical control activity but is not present with each gastric slow wave. Fast and slow acting calcium channels lead to different strengths of contractions, where fast calcium channels have been associated with stronger contractions (>0.5 gf) and slow calcium channels have been associated with weaker contractions (<0.5gf). The transmembrane activity of all the ionic channels can be detected and monitored using intracellular recordings [160], however the electrical impact of the calcium channels is of lower average electrical power, and dissipates in the body before reaching the skin. A number of factors impact how electricity travels through the body, including body mass index and tissue impedances. As a result, the activity of the calcium channels cannot reliably contribute to the power dynamics of cutaneously recorded EGG. While the present study aimed to further validate the TIIM technique, the sham pill study was effectively a direct comparison between the gastric motility indices obtained via routine EGG and force transducers, since standard EGG filtering parameters were used. In four animals statistically significant correlation was not observed between any of the channel pairs examined, further supporting our understanding that EGG cannot reliably monitor gastric motility.

Conversely, the electrical signal emitted by a man-made oscillator introduced into the lumen of the stomach can reach the abdominal skin with reliable and easily detectable parameters. The implementation of the TIIM technique was to utilize electrical power and frequency factors that would minimize inconsistent dissipation within the body while exhibiting a negligible effect on all tissue types. This signal is subjected to attenuation throughout each layer of tissue as it travels to the outer layer of the skin, and small changes to this attenuation due to dynamic tissues can be recorded cutaneously as evoked skin potentials. The TIIM signal is much stronger than the induced electrical phenomena of the fast and slow calcium channels, but is still affected by the mechanical activity they represent. In the case of TIIM, the most dynamic attenuation was the result of contracting gastric muscle in the vicinity of the retainable oscillator, which appeared as an amplitude modulated signal.

While the present study was focused on demonstrating the effectiveness of a gastric retentive pill-based implementation of the TIIM technique, the concept itself had been proved earlier using a catheter based oscillator. This catheter oscillator required intubation, which for acute stationary animal studies is not an issue, but is uncomfortable and inconvenient for ambulatory human tests, which provided the motivation to move ahead with a pill-based design. One distinct advantage of the catheter design was that because the oscillator circuit itself was external to the body, it could share an electrical ground with the custom EGG used for the feasibility study. By sharing a common reference voltage, the measurements become more reliable, and unwanted noise from electrically dynamic environments can be reduced. The catheter was also able to provide more information, such as the pH detected at the tip of the catheter. The benefit of this was the added position verification and monitoring; if a sudden change in pH occurred it could indicate a shift in the catheter. While existing pill-based technology has pH monitoring capabilities, it was not the focus of the present study, however it could be added to future iterations to create a solution similar to the wireless motility capsule. The pill-based TIIM pill is much more convenient to consume, and has the added benefit of the gastric retentive enclosure. The expandable and compressible polymers around the pill most likely aided the TIIM signal by ensuring consistent contact with the walls of the lumen, making

the results less subject to whether or not the electrodes were in contact or what media the signal travelled through, as would have been in the case of the catheter design. Depending on future ambulatory studies it could appear that the catheter design offers far superior EMR noise reduction, but it could be that the convenience of the pill makes it more appealing, similar to the wireless motility capsule or capsule endoscopy.

With the development of the portable TIIM receiver it is now possible to conduct long- term ambulatory chronic tests. Longer tests could examine more clinically relevant cases such as drinking and eating, or even just the confirmation of the migrating motor complex. During meals it is likely that the impedance of the ingested content would influence the TIIM signal. This may appear as a slow change in the baseline impedance, over which the gastric contractions are imposed. If this were the case it should be validated against gastric emptying scintigraphy, and could position TIIM as an all-encompassing test in which muscle tone and gastric emptying can be monitored simultaneously. Furthermore, if the TIIM capsule were set to turn off periodically, the inherent electrical activity of the stomach would be recorded, providing every possible metric to a physician looking for clues to guide therapy.

All of the technology required to move the TIIM technique to market and practical clinical use exists today, and most is already being utilized for other GI studies. Most notably of these is the body conduction technique developed by Intromedic, which transmits digital information through the tissue at approximately 300 kHz, still within the range of low frequency dynamics. While the receiver for the digital video from the capsule endoscope does not analyze the attenuation of the signal, it would not be difficult to perform a similar study to the one presented in this thesis using a capsule endoscope instead of the TIIM transducer pill. A capsule endoscope could be encased in the same gastric retentive technology utilized within this study, and the attenuation dynamics could be monitored using an available EGG. If it was further combined with the technology available in the wireless motility capsule, a full GI transit study could be performed after the gastric retentive mesh has disintegrated. In either case, this could

provide a unique opportunity to revive the applicability of the EGG using available and in some cases coded technologies. By using technologies that have already been proven safe and reliable, precedent for human trials could be established in order to investigate the relationship of TIIM signals to disease states.

Although the reported results of this new technology are encouraging, further testing and development is required before moving to human testing. It is possible that environmental phenomena such as capsule tumbling due to motion not associated with gastric contractions, static electricity and external electromagnetic radiation may affect the measurements in an ambulatory setting, particularly if the patient is mobile in an electromagnetically-dynamic environment. In addition, it is possible that abdominal muscle contractions and motion artifacts could be registered. Adaptive filter procedures might help to eliminate unwanted artifacts [161], but at the price of increasing complexity.

One possibility to reduce the impact of both physiological and mechanical motion artifacts could be to implement accelerometers into the pill and receiver. The data from the accelerometer could be transmitted digitally to the receiver using body conduction technology similar to that used in capsule endoscopy, and a simple digital filter could be implemented to subtract the power of the pill motion from the power of the accelerometer in the receiver. In this way erroneous data from motion linked to walking or shifting body position can be filtered out, enabling higher quality recordings.

The TIIM technique further offers the possibility to interrogate other organs in the vicinity of the stomach, including most notably the liver. Placing cutaneous electrodes over the liver could offer insight into the stiffness of the organ, which has been implicated in disease states [162]. In this case TIIM does not have to be limited to an electrical signal; by exchanging

the electrical transducer with a sonic or ultrasonic transducer acoustic impedance of the tissue could be assessed. Figure 4-1 shows a proposed method to assess the impedance of the liver.

Figure 4-1: Proposed technique for measuring the impedance of the liver. In this case the internal transducer may be acoustic or electrical, depending on the desired results.

Chapter 5: Conclusion Transcutaneous Intraluminal Impedance Measurement (TIIM) is a minimally invasive gastric motility monitoring technique which has now been validated in a sham-comparative study. A fully designed pill-based oscillator was described and implemented for use in a study to compare the effectiveness of TIIM and invasively implanted force transducers in acute canines. Calculating one-minute gastric motility indices showed that TIIM was able to measure gastric motility with statistically significant correlation compared to the force transducers (p<0.01), which was not observed in the sham studies (p>0.1). Further spectral analysis of the signals revealed that TIIM demonstrated contractions per minute with significant correlation to the force transducers (p<0.05), which was also not observed in the sham studies (p>0.1). TIIM can be regarded as a possible replacement test by which to analyze and observe gastric motility.

Furthermore, a novel portable TIIM receiver has been designed constructed and tested. The data recorded from the TIIM receiver was significantly correlated to the input signal, which was created from an original recording taken using equipment utilized in the animal studies of this Thesis (p<0.01). Further improvements and testing are required before implementing the technology.

Chapter 6: Scholarly Contributions Journal Articles:

Poscente, M., Wang, G., Filip, D., Ninova, P., Yadid-Pecht, O., Andrews, C. N., and Mintchev, M. P., Real-time gastric motility monitoring using transcutaneous intraluminal impedance measurements (TIIM). Physiological Measurement, 2014. 35(2): p. 217.

Poscente, M., Wang, G., Filip, D., Ninova, P., Muench, G., Yadid-Pecht, O., Mintchev, M. P., and Andrews, C. N., Transcutaneous intraluminal impedance measurement for minimally invasive monitoring of gastric motility: validation in acute canine models. Gastroengerology Research and Practice, 2014. 2014.

Wang, G., Poscente, M. D., Park, S. S., Yadid-Pecht, O., and Mintchev, M. P., Characterization of a cuff-based shape memory alloy (SMA) actuator. International Journal “Information Theories and Applications”, 2015. 22(1): p. 3.

Wang, Z., Poscente, M., Filip, D., Dimanchev, M., and Mintchev, M., Rotary in-drilling alignment using an autonomous MEMS-based inertial measurement unit for measurement-while-drilling processes. Instrumentation & Measurement Magazine, IEEE, 2013. 16(6): p. 26-34.

Conference Presentations:

Poscente, M. D., Hussain, A., Filip, D., Andrews, C. N., & Mintchev, M. P., Mo2086 Transcutaneous Intraluminal Impedance Measurements (TIIM): A New Minimally-Invasive Technique for Long-Term Monitoring of Gastric Motility. Gastroenterology, 2013. 144(5), S-737.

Wang, G., Poscente, M. D., Filip, D., Yadid-Pecht, O., Andrews, C. N., & Mintchev, M. P., Mo1302 gastric-retentive transcutaneous intraluminal impedance measurement (TIIM): sham controlled, minimally-invasive assessment of gastric motility in acute canine models. Gastroenterology, 2014. 146(5), S-613.

Patent Applications:

Poscente, M. D., Yadid-Pecht, O., Andrews, C. N., Mintchev, M. P., “Device and method for monitoring internal organs.” U.S. Patent Application 14/914,397, Filed Feb. 25, 2016.

References 1. Kim, T.H., and Shivdasani, R.A., Stomach development, stem cells and disease. Development, 2016. 143(4): p. 554-565. 2. Kelly, K.A., Gastric emptying of liquids and solids: roles of proximal and distal stomach. American Journal of Physiology-Gastrointestinal and Liver Physiology, 1980. 239(2): p. G71-G76. 3. Sarna, S.K., Gastrointestinal longitudinal muscle contractions. American Journal of Physiology-Gastrointestinal and Liver Physiology, 1993. 265(1): p. G156-G164. 4. McSwiney, B., Innervation of the stomach. Physiological Reviews, 1931. 11(4): p. 478- 514. 5. Janssen, P., et al., Review article: the role of gastric motility in the control of food intake. Alimentary Pharmacology & Therapeutics, 2011. 33(8): p. 880-894. 6. Furness, J.B., and Costa, M., The enteric nervous system. 1987: Churchill Livingstone Edinburgh. 7. Debas, H.T., Farooq, O., and Grossman, M.I., Inhibition of gastric emptying is a physiological action of cholecystokinin. Gastroenterology, 1975. 68(5 Pt 1): p. 1211- 1217. 8. Levein, N., Thörn, S., and Wattwil, M., Dopamine delays gastric emptying and prolongs orocaecal transit time in volunteers. European Journal of Anaesthesiology, 1999. 16(4): p. 246-250. 9. Camilleri, M., et al., Measurement of gastrointestinal motility in the GI laboratory. Gastroenterology, 1998. 115(3): p. 747-762. 10. Hughes, M.J., Andrews, D.W., and Huizinga, J.D., Interstitial cells of Cajal generate a rhythmic pacemaker current. Nature Medicine, 1998. 4(7). 11. Yun, H.Y., et al., Regional distribution of interstitial cells of cajal (ICC) in human stomach. The Korean Journal of Physiology & Pharmacology, 2010. 14(5): p. 317-324. 12. Vantrappen, G., et al., Motilin and the interdigestive migrating motor complex in man. Digestive Diseases and Sciences, 1979. 24(7): p. 497-500.

13. Sanger, G.J., Broad, J., and Andrews, P.L., The relationship between gastric motility and nausea: gastric prokinetic agents as treatments. European Journal of Pharmacology, 2013. 715(1): p. 10-14. 14. Bouras, E.P., Roque, M.I.V., and Aranda-Michel, J., Gastroparesis From Concepts to Management. Nutrition in Clinical Practice, 2013. 28(4): p. 437-447. 15. Livingston, E.H., and Passaro JR, E.P., Postoperative ileus. Digestive Diseases and Sciences, 1990. 35(1): p. 121-132. 16. Kao, C.H., et al., Gastric emptying in spinal cord injury patients. Digestive Diseases and Sciences, 1999. 44(8): p. 1512-1515. 17. Horowitz, M., et al., Gastric emptying in diabetes: an overview. Diabetic Medicine: A Journal of the British Diabetic Association, 1996. 13(9 Suppl 5): p. S16-22. 18. Marathe, C.S., et al., Novel insights into the effects of diabetes on gastric motility. Expert Review of Gastroenterology & Hepatology, 2016. 19. Wilson, V.L., Gastroparesis: a patient's experience. Journal of Diabetes Nursing, 2004. 8(2): p. 73-75. 20. Ordög, T., Interstitial cells of Cajal in diabetic gastroenteropathy. Neurogastroenterology and Motility: The Official Journal of the European Gastrointestinal Motility Society, 2008. 20(1): p. 8-18. 21. Lacy, B., et al., Review article: current treatment options and management of functional dyspepsia. Alimentary Pharmacology & Therapeutics, 2012. 36(1): p. 3-15. 22. Talley, N.J., and Ford, A.C., Functional dyspepsia. New England Journal of Medicine, 2015. 373(19): p. 1853-1863. 23. Drossman, D.A., The functional gastrointestinal disorders and the Rome III process. Gastroenterology, 2006. 130(5): p. 1377-1390. 24. Bramble, M., Suvakovic, Z., and Hungin, A., Detection of upper gastrointestinal cancer in patients taking antisecretory therapy prior to gastroscopy. Gut, 2000. 46(4): p. 464-467. 25. Tack, J., and Talley, N.J., Functional dyspepsia—symptoms, definitions and validity of the Rome III criteria. Nature Reviews Gastroenterology and Hepatology, 2013. 10(3): p. 134- 141.

26. Dibaise, J., et al., Psychological distress in Rome III functional dyspepsia patients presenting for testing of gastric emptying. Neurogastroenterology & Motility, 2015. 27. Camilleri, M., and Stanghellini, V., Current management strategies and emerging treatments for functional dyspepsia. Nature Reviews Gastroenterology and Hepatology, 2013. 10(3): p. 187-194. 28. Barkun, M., et al., Evidence-based recommendations for short-and long-term management of uninvestigated dyspepsia in primary care: an update of the Canadian Dyspepsia Working Group (CanDys) clinical management tool. Canadian Journal of Gastroenterology, 2005. 19(5): p. 285. 29. Ford, A.C., et al., Helicobacter pylori “test and treat” or endoscopy for managing dyspepsia: an individual patient data meta-analysis. Gastroenterology, 2005. 128(7): p. 1838-1844. 30. Delaney, B., Moayyedi, P., and Forman, D., Initial management strategies for dyspepsia. The Cochrane Library, 2003. 31. Lassen, A., Hallas, J., and de Muckadell, O.S., Helicobacter pylori test and eradicate versus prompt endoscopy for management of dyspeptic patients: 6.7 year follow up of a randomised trial. Gut, 2004. 53(12): p. 1758-1763. 32. Rabeneck, L., et al., A double blind, randomized, placebo-controlled trial of proton pump inhibitor therapy in patients with uninvestigated dyspepsia. The American Journal of Gastroenterology, 2002. 97(12): p. 3045-3051. 33. Spiegel, B.M., Vakil, N.B., and Ofman, J.J., Dyspepsia management in primary care: a decision analysis of competing strategies. Gastroenterology, 2002. 122(5): p. 1270-1285. 34. Futagami, S., et al., Nizatidine improves clinical symptoms and gastric emptying in patients with functional dyspepsia accompanied by impaired gastric emptying. Digestion, 2012. 86(2): p. 114-121. 35. Xiao, Y. L., et al., Prevalence and symptom pattern of pathologic esophageal acid reflux in patients with functional dyspepsia based on the Rome III criteria. The American Journal of Gastroenterology, 2010. 105(12): p. 2626-2631.

36. Moayyedi, P., et al., Pharmacological interventions for non-ulcer dyspepsia. Cochrane Database System Review, 2006. 4. 37. Camilleri, M., Functional dyspepsia: mechanisms of symptom generation and appropriate management of patients. Gastroenterology Clinics of North America, 2007. 36(3): p. 649-664. 38. Tack, J., et al., Influence of the selective serotonin re‐uptake inhibitor, paroxetine, on gastric sensorimotor function in humans. Alimentary Pharmacology & Therapeutics, 2003. 17(4): p. 603-608. 39. Dinan, T., et al., A double‐blind placebo‐controlled study of buspirone‐stimulated prolactin release in non‐ulcer dyspepsia—are central serotoninergic responses enhanced? Alimentary Pharmacology & Therapeutics, 2001. 15(10): p. 1613-1618. 40. Camilleri, M., et al., Clinical guideline: management of gastroparesis. The American Journal of Gastroenterology, 2013. 108(1): p. 18-37. 41. Sarnelli, G., et al., Symptoms associated with impaired gastric emptying of solids and liquids in functional dyspepsia. The American Journal of Gastroenterology, 2003. 98(4): p. 783-788. 42. Tack, J., Gastric motor disorders. Best Practice & Research Clinical Gastroenterology, 2007. 21(4): p. 633-644. 43. Casaer, M.P., et al., Early versus late parenteral nutrition in critically ill adults. New England Journal of Medicine, 2011. 365(6): p. 506-517. 44. Roberts, P., Nutrition in the head-injured patient. New Horizons (Baltimore, Md.), 1995. 3(3): p. 506-517. 45. Locke, G.R., et al., Risk of gastroparesis in subjects with type 1 and 2 diabetes in the general population. The American Journal of Gastroenterology, 2012. 107(1): p. 82-88. 46. Talley, N.J., et al., Impact of chronic gastrointestinal symptoms in diabetes mellitus on health-related quality of life. The American Journal of Gastroenterology, 2001. 96(1): p. 71-76.

47. Bharucha, A.E., et al., Relationship between glycemic control and gastric emptying in poorly controlled type 2 diabetes. Clinical Gastroenterology and Hepatology, 2015. 13(3): p. 466-476. e1. 48. Olausson, E.A., et al., Measurement of gastric emptying by radiopaque markers in patients with diabetes: correlation with scintigraphy and upper gastrointestinal symptoms. Neurogastroenterology & Motility, 2013. 25(3): p. e224-e232. 49. Punkkinen, J., et al., Upper abdominal symptoms in patients with Type 1 diabetes: unrelated to impairment in gastric emptying caused by autonomic neuropathy. Diabetic Medicine, 2008. 25(5): p. 570-577. 50. Schvarcz, E., et al., Physiological hyperglycemia slows gastric emptying in normal subjects and patients with insulin-dependent diabetes mellitus. Gastroenterology, 1997. 113(1): p. 60-66. 51. Richards, R.D., Davenport, K., and McCallum, R.W., The treatment of idiopathic and diabetic gastroparesis with acute intravenous and chronic oral erythromycin. American Journal of Gastroenterology, 1993. 88(2). 52. Buckler, K., Effects of gastric surgery upon gastric emptying in cases of peptic ulceration. Gut, 1967. 8(2): p. 137-147. 53. Barnes, A., and Williams, J.A., Stomach drainage after vagotomy and pyloroplasty. The American Journal of Surgery, 1967. 113(4): p. 494-497. 54. Parkman, H.P., et al., Similarities and differences between diabetic and idiopathic gastroparesis. Clinical Gastroenterology and Hepatology, 2011. 9(12): p. 1056-1064. 55. Camilleri, M., Appraisal of medium-and long-term treatment of gastroparesis and chronic intestinal dysmotility. American Journal of Gastroenterology, 1994. 89(10). 56. Parkman, H.P., et al., Dietary intake and nutritional deficiencies in patients with diabetic or idiopathic gastroparesis. Gastroenterology, 2011. 141(2): p. 486-498. e7. 57. Low, A., Nutritional regulation of gastric secretion, digestion and emptying. Nutrition Research Reviews, 1990. 3(1): p. 229-252. 58. Howarth, N.C., Saltzman, E., and Roberts, S.B., Dietary fiber and weight regulation. Nutrition Reviews, 2001. 59(5): p. 129-139.

59. McCance, D., Dietary fiber and gastrointestinal function. 1987. 60. Ogorek, C.P., et al., Idiopathic gastroparesis is associated with a multiplicity of severe dietary deficiencies. American Journal of Gastroenterology, 1991. 86(4). 61. Devlin, M.J., et al., Gastric emptying and symptoms of bulimia nervosa: effect of a prokinetic agent. Physiology & Behavior, 2012. 106(2): p. 238-242. 62. Acosta, A., and Camilleri, M., Prokinetics in gastroparesis. Gastroenterology Clinics of North America, 2015. 44(1): p. 97-111. 63. Tack, J., Carbone, F., and Rotondo, A., Gastroparesis. Current Opinion in Gastroenterology, 2015. 31(6): p. 499-505. 64. Gillis, M., and Koch, K.L., Gastroparesis: Pathophysiology of Chronic Abdominal Pain and Current Treatment, in Chronic Abdominal Pain. 2015, Springer. p. 69-75. 65. McCallum, R., Sarosiek, I., and Bashashati, M., Safety of treatment for gastroparesis. Expert Opinion on Drug Safety, 2016. 66. Stevens, J.E., et al., Pathophysiology and pharmacotherapy of gastroparesis: current and future perspectives. Expert Opinion on Pharmacotherapy, 2013. 14(9): p. 1171-1186. 67. Hasler, W.L., Symptomatic Management for Gastroparesis: Antiemetics, Analgesics, and Symptom Modulators. Gastroenterology Clinics of North America, 2015. 44(1): p. 113- 126. 68. Roth, A.J., and Alici, Y., General principles of psychopharmacological treatment in psycho-oncology, in Psychopharmacology in Oncology and Palliative Care. 2014, Springer. p. 13-30. 69. Liu, Y., et al., The effects of comprehensive mental intervention on the recovery time of patients with postsurgical gastroparesis syndrome. Journal of Clinical Nursing, 2014. 23(21-22): p. 3138-3147. 70. Parrish, C.R., Nutritional Considerations in the Patient with Gastroparesis. Gastroenterology Clinics of North America, 2015. 44(1): p. 83-95. 71. Roses, A.P.R.E., and Fraker, D.L., Bile Reflux and Gastroparesis, in Gastrointestinal Surgery. 2015, Springer. p. 119-125.

72. Yang, D.D., et al., Risk factors of gastroparesis syndrome after abdominal non- gastroduodenal operation and its prevention. Asian Pacific Journal of Tropical Medicine, 2013. 6(6): p. 497-499. 73. Rodriguez, L., et al., Endoscopic intrapyloric injection of botulinum toxin A in the treatment of children with gastroparesis: a retrospective, open-label study. Gastrointestinal Endoscopy, 2012. 75(2): p. 302-309. 74. Ukleja, A., et al., Endoscopic botox injections in therapy of refractory gastroparesis. World Journal of Gastrointestinal Endoscopy, 2015. 7(8): p. 790. 75. Heckert, J., et al., Gastric Electric Stimulation for Refractory Gastroparesis: A Prospective Analysis of 151 Patients at a Single Center. Digestive Diseases and Sciences, 2016. 61(1): p. 168-175. 76. Zehetner, J., et al., Minimally invasive surgical approach for the treatment of gastroparesis. Surgical Endoscopy, 2013. 27(1): p. 61-66. 77. Abell, T., et al., Gastric electrical stimulation for medically refractory gastroparesis. Gastroenterology, 2003. 125(2): p. 421-428. 78. McCallum, R.W., et al., Gastric electrical stimulation with Enterra therapy improves symptoms from diabetic gastroparesis in a prospective study. Clinical Gastroenterology and Hepatology, 2010. 8(11): p. 947-954. e1. 79. Kordzadeh, A., et al., OC-086 Gastric electrical stimulation (ges) for the treatment of refractory gastroparesis: a prospective 14-year experience in a UK centre. Gut, 2015. 64(Suppl 1): p. A43-A43. 80. Keller, D.S., et al., Surgical outcomes after gastric electric stimulator placement for refractory gastroparesis. Journal of Gastrointestinal Surgery, 2013. 17(4): p. 620-626. 81. Toro, J.P., et al., Efficacy of laparoscopic pyloroplasty for the treatment of gastroparesis. Journal of the American College of Surgeons, 2014. 218(4): p. 652-660. 82. Stanghellini, V., and Tack, J., Gastroparesis: separate entity or just a part of dyspepsia? Gut, 2014. 63(12): p. 1972-1978. 83. Saltrelli, C.M., Living With Gastroparesis: A Little Hope Goes a Long Way. The American Journal of Gastroenterology, 2016.

84. Forgacs, I., and Raja, O., Diabetes and the gastrointestinal tract. Medicine, 2015. 43(6): p. 347-351. 85. Szarka, L.A., and Camilleri, M., Methods for measurement of gastric motility. American Journal of Physiology-Gastrointestinal and Liver Physiology, 2009. 296(3): p. G461-G475. 86. Akkermans, L., and Van Isselt, J., Gastric motility and emptying studies with radionuclides in research and clinical settings. Digestive Diseases and Sciences, 1994. 39(12): p. 95S-96S. 87. Camilleri, M., et al., Towards a less costly but accurate test of gastric emptying and small bowel transit. Digestive Diseases and Sciences, 1991. 36(5): p. 609-615. 88. Abell, T.L., et al., Consensus recommendations for gastric emptying scintigraphy: a joint report of the American Neurogastroenterology and Motility Society and the Society of Nuclear Medicine. The American Journal of Gastroenterology, 2008. 103(3): p. 753-763. 89. Macfie, A., et al., Gastric emptying in pregnancy. British Journal of Anaesthesia, 1991. 67(1): p. 54-57. 90. Feinle, C., et al., Scintigraphic validation of a magnetic resonance imaging method to study gastric emptying of a solid meal in humans. Gut, 1999. 44(1): p. 106-111. 91. Camilleri, M., et al., Performance characteristics of scintigraphic measurement of gastric emptying of solids in healthy participants. Neurogastroenterology & Motility, 2012. 24(12): p. 1076-e562. 92. Degen, L., and Phillips, S., Variability of gastrointestinal transit in healthy women and men. Gut, 1996. 39(2): p. 299-305. 93. Camilleri, M., et al., Relation between antral motility and gastric emptying of solids and liquids in humans. American Journal of Physiology-Gastrointestinal and Liver Physiology, 1985. 249(5): p. G580-G585. 94. Bonta, D.V., et al., Clinical Intervention for Quality Improvement of Gastric-Emptying Studies. Journal of Nuclear Medicine Technology, 2014. 42(4): p. 274-277. 95. Viramontes, B., et al., Validation of a stable isotope gastric emptying test for normal, accelerated or delayed gastric emptying. Neurogastroenterology & Motility, 2001. 13(6): p. 567-574.

96. Braden, B., et al., The [13 C] acetate breath test accurately reflects gastric emptying of liquids in both liquid and semisolid test meals. Gastroenterology, 1995. 108(4): p. 1048- 1055. 97. Nguyen, N.Q., et al., Gastric emptying measurement of liquid nutrients using the 13C- octanoate breath test in critically ill patients: a comparison with scintigraphy. Intensive Care Medicine, 2013. 39(7): p. 1238-1246. 98. Dickman, R., et al., Comparison of continuous breath test and gastric scintigraphy for the measurement of gastric emptying rate in healthy and dyspeptic individuals. European Journal of Gastroenterology & Hepatology, 2013. 25(3): p. 291-295. 99. Bharucha, A.E., et al., Comprehensive assessment of gastric emptying with a stable isotope breath test. Neurogastroenterology & Motility, 2013. 25(1): p. e60-e69. 100. Sadik, R., Abrahamsson, H., and Stotzer, P., Gender differences in gut transit shown with a newly developed radiological procedure. Scandinavian Journal of Gastroenterology, 2003. 38(1): p. 36-42. 101. Horowitz, M., et al., Changes in gastric emptying rates with age. Clinical Science, 1984. 67(2): p. 213-218. 102. Rune, S., and Viskum, K., Duodenal pH values in normal controls and in patients with duodenal ulcer. Gut, 1969. 10(7): p. 569-571. 103. Byrne, K.G., and Quigley, E., Antroduodenal manometry: an evaluation of an emerging methodology. Digestive Diseases, 1997. 15(Suppl. 1): p. 53-63. 104. Jebbink, R.J., et al., Antroduodenal manometry: 24-hour ambulatory monitoring versus short-term stationary manometry in patients with functional dyspepsia. European Journal of Gastroenterology & Hepatology, 1995. 7(2): p. 109-116. 105. Desipio, J., et al., High‐resolution solid‐state manometry of the antropyloroduodenal region. Neurogastroenterology & Motility, 2007. 19(3): p. 188-195. 106. Frank, J.W., Sarr, M.G., and Camilleri, M., Use of gastroduodenal manometry to differentiate mechanical and functional intestinal obstruction: an analysis of clinical outcome. American Journal of Gastroenterology, 1994. 89(3).

107. Holland, R., Gallagher, M. D., and Quigley, E. M., An evaluation of an ambulatory manometry system in assessment of antroduodenal motor activity. Digestive Diseases and Sciences, 1996. 41(8): p. 1531-1537. 108. J. van der Schaar, C.L., and Masclee, P., The role of the barostat in human research and clinical practice. Scandinavian Journal of Gastroenterology, 1999. 34(230): p. 52-63. 109. De Schepper, H., et al., Assessment of gastric accommodation: overview and evaluation of current methods. Neurogastroenterology & Motility, 2004. 16(3): p. 275-285. 110. Sauter, M., et al., Toward more accurate measurements of anorectal motor and sensory function in routine clinical practice: Validation of High‐Resolution Anorectal Manometry and Rapid Barostat Bag measurements of rectal function. Neurogastroenterology & Motility, 2014. 26(5): p. 685-695. 111. Pfaffenbach, B., et al., Electrogastrography in healthy subjects. Digestive Diseases and Sciences, 1995. 40(7): p. 1445-1450. 112. Smout, A., Van der Schee, E.J., and Grashuis, J., What is measured in electrogastrography? Digestive Diseases and Sciences, 1980. 25(3): p. 179-187. 113. Van der Voort, I., et al., Electrogastrography as a diagnostic tool for delayed gastric emptying in functional dyspepsia and irritable bowel syndrome. Neurogastroenterology & Motility, 2003. 15(5): p. 467-473. 114. Stern, R. M., et al., Electrogastrography: current issues in validation and methodology. Psychophysiology, 1987. 24(1): p. 55-64. 115. Kloetzer, L., et al., Motility of the antroduodenum in healthy and gastroparetics characterized by wireless motility capsule. Neurogastroenterology & Motility, 2010. 22(5): p. 527-e117. 116. Maqbool, S., Parkman, H.P., and Friedenberg, F.K., Wireless capsule motility: comparison of the SmartPill® GI monitoring system with scintigraphy for measuring whole gut transit. Digestive Diseases and Sciences, 2009. 54(10): p. 2167-2174. 117. Tran, K., Brun, R., and Kuo, B., Evaluation of regional and whole gut motility using the wireless motility capsule: relevance in clinical practice. Therapeutic Advances in Gastroenterology, 2012. 5(4): p. 249-260.

118. Cassilly, D., et al., Gastric emptying of a non‐digestible solid: assessment with simultaneous SmartPill pH and pressure capsule, antroduodenal manometry, gastric emptying scintigraphy. Neurogastroenterology & Motility, 2008. 20(4): p. 311-319. 119. Hasler, W.L., The use of SmartPill for gastric monitoring. Expert Review of Gastroenterology & Hepatology, 2014. 8(6): p. 587-600. 120. Marciani, L., et al., Effect of meal viscosity and nutrients on satiety, intragastric dilution, and emptying assessed by MRI. American Journal of Physiology-Gastrointestinal and Liver Physiology, 2001. 280(6): p. G1227-G1233. 121. Marciani, L., Assessment of gastrointestinal motor functions by MRI: a comprehensive review. Neurogastroenterology & Motility, 2011. 23(5): p. 399-407. 122. Couturier, D., et al., Electrical activity of the normal human stomach. The American Journal of Digestive Diseases, 1972. 17(11): p. 969-976. 123. Yang, W., and Peng, L., Image reconstruction algorithms for electrical capacitance tomography. Measurement Science and Technology, 2002. 14(1): p. R1. 124. Borcea, L., Electrical impedance tomography. Inverse Problems, 2002. 18(6): p. R99. 125. Bodenstein, M., David, M., and Markstaller, K., Principles of electrical impedance tomography and its clinical application. Critical Care Medicine, 2009. 37(2): p. 713-724. 126. Arriagada, A., et al., Design, implementation and testing of an implantable impedance- based feedback-controlled neural gastric stimulator. Physiological Measurement, 2011. 32(8): p. 1103. 127. Zarowitz, B., Peterson, E., and Robert, S., Characterization of drug disposition and dosing using bioelectrical impedance. Medical Progress Through Technology, 1992. 19(4): p. 193-198. 128. Shim, H.B., et al., Endoscope and a method for finding its location. 2006, U.S. Patent Application 12/442,836, filed September 28, 2006. 129. Poscente, M., et al., Real-time gastric motility monitoring using transcutaneous intraluminal impedance measurements (TIIM). Physiological Measurement, 2014. 35(2): p. 217.

130. Stemper, T., Gastric emptying and its relationship to antral contractile activity. Gastroenterology, 1975. 69(3): p. 649-653. 131. Snedecor, G., and Cochran, W., Statistical Methods. 1967, The Iowa State University Press, Ames, IA. 132. Toennies, J.L., et al., Swallowable medical devices for diagnosis and surgery: the state of the art. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 2010. 224(7): p. 1397-1414. 133. Scrosati, B., Croce, F., and Panero, S., Progress in lithium polymer battery R&D. Journal of Power Sources, 2001. 100(1): p. 93-100. 134. Tarascon, J.M., and Armand, M., Issues and challenges facing rechargeable lithium batteries. Nature, 2001. 414(6861): p. 359-367. 135. Peluso, G., et al., Polyetherimide as biomaterial: preliminary in vitro and in vivo biocompatibility testing. Journal of Materials Science: Materials in Medicine, 1994. 5(9- 10): p. 738-742. 136. Protection, I.C.o.N.-I.R., Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz to 100 kHz). Health Physics, 2010. 99(6): p. 818-836. 137. Liu, M., and Guo, T., Preparation and swelling properties of crosslinked sodium polyacrylate. Journal of Applied Polymer Science, 2001. 82(6): p. 1515-1520. 138. Tack, J., et al., Role of impaired gastric accommodation to a meal in functional dyspepsia. Gastroenterology, 1998. 115(6): p. 1346-1352. 139. Deneva, M.G., et al., Utilization of temporary controllable intragastric pseudobezoars for the treatment of obesity. Current Obesity Reports, 2012. 1(2): p. 68-74. 140. Dive, A., et al., Effect of erythromycin on gastric motility in mechanically ventilated critically ill patients: a double-blind, randomized, placebo-controlled study. Critical Care Medicine, 1995. 23(8): p. 1356-1362. 141. Sedra, A.S., and Smith, K.C., Microelectronic circuits. Vol. 1. 1998: New York: Oxford University Press. 142. Hall, J.A., et al., Gastric myoelectric and motor activity in dogs after isoflurane anesthesia. Veterinary Surgery, 1995. 24(5): p. 456-463.

143. Reinke, D.A., Rosenbaum, A.H., and Bennett, D.R., Patterns of dog gastrointestinal contractile activity monitored in vivo with extraluminal force transducers. Digestive Diseases and Sciences, 1967. 12(2): p. 113-141. 144. Poscente, M.D., et al., Transcutaneous Intraluminal Impedance Measurement for Minimally Invasive Monitoring of Gastric Motility: Validation in Acute Canine Models. Gastroenterology Research and Practice, 2014. 2014. 145. Kee, W.C., et al., Optimal placement of impedance epigastrography electrodes. Annals of Biomedical Engineering, 1996. 24(2): p. 328-332. 146. Kelly, K., Code, C.F., and Elveback, L.R., Patterns of canine gastric electrical activity. American Journal of Physiology, 1969. 217(2): p. 461-470. 147. Guariguata, L., et al., Global estimates of diabetes prevalence for 2013 and projections for 2035. Diabetes Research and Clinical Practice, 2014. 103(2): p. 137-149. 148. Nanditha, A., et al., Diabetes in Asia and the Pacific: Implications for the Global Epidemic. Diabetes Care, 2016. 39(3): p. 472-485. 149. Zarrinpar, A., and Loomba, R., Review article: the emerging interplay among the gastrointestinal tract, bile acids and incretins in the pathogenesis of diabetes and non‐ alcoholic fatty liver disease. Alimentary Pharmacology & Therapeutics, 2012. 36(10): p. 909-921. 150. Bharucha, A.E., et al., Delayed gastric emptying is associated with early and long-term hyperglycemia in type 1 diabetes mellitus. Gastroenterology, 2015. 149(2): p. 330-339. 151. Phillips, L.K., et al., Measurement of gastric emptying in diabetes. Journal of Diabetes and its Complications, 2014. 28(6): p. 894-903. 152. Patcharatrakul, T., and Gonlachanvit, S., Technique of functional and motility test: how to perform antroduodenal manometry. Journal of Neurogastroenterology and Motility, 2013. 19(3): p. 395-404. 153. Arora, Z., et al., Clinical utility of wireless motility capsule in patients with suspected multiregional gastrointestinal dysmotility. Digestive Diseases and Sciences, 2015. 60(5): p. 1350-1357.

154. Saad, R.J., The Wireless Motility Capsule: a One-Stop Shop for the Evaluation of GI Motility Disorders. Current Gastroenterology Reports, 2016. 18(3): p. 1-7. 155. Leung, W., et al., Alternatives to colonoscopy for population-wide colorectal cancer screening. Hong Kong Medical Journal, 2016: p. 22. 156. Smout, A., et al., Role of electrogastrography and gastric impedance measurements in evaluation of gastric emptying and motility. Digestive Diseases and Sciences, 1994. 39(12): p. 110S-113S. 157. Bredenoord, A.J., and Smout, A.J., Advances in motility testing—current and novel approaches. Nature Reviews Gastroenterology and Hepatology, 2013. 10(8): p. 463-472. 158. Sarna, S., and Otterson, M., Gastrointestinal motility: some basic concepts. Pharmacology, 1988. 36(Suppl. 1): p. 7-14. 159. Szurszewski, J.H., Electrical basis for gastrointestinal motility. Physiology of the Gastrointestinal Tract, 1987. 2: p. 383-422. 160. Smout, A.J.P.M., Myoelectric activity of the stomach: gastroelectromyography and electrogastrography. 1980. 161. Adochiei, N.I., David, V., and Tudosa, I., Methods of electromagnetic interference reducion in electrocardiographic signal acquisiiton. Environmental Engineering & Management Journal (EEMJ), 2011. 10(4). 162. Marcellin, P., et al., Non‐invasive assessment of liver fibrosis by stiffness measurement in patients with chronic hepatitis B. Liver International, 2009. 29(2): p. 242-247.

Appendices APPENDIX A: Correlation analysis and one-minute motility index code

LMinute = 30; Fs=10; %Sampling frequency; NoChl=4; %No. of Channels;

Time1=480:(1/Fs):((LMinute*60)-(1/Fs)); data1(:,1)=Channel3; data1(:,2)=Channel3; data1(:,3)=Channel4; data1(:,4)=Channel5; close all; figure(1); subplot(8,1,1); % plot(Time1,data1(1:(LMinute*60*10),1),'r-'); % title('The Raw Data: Channel 1: EGG 1'); subplot(8,1,2); % plot(Time1,data1(1:(LMinute*60*10),2),'r-'); % title('The Raw Data: Channel 2: EGG 2'); subplot(8,1,3); % plot(Time1,data1(1:(LMinute*60*10),3),'r-'); % title('The Raw Data: Channel 3: Force Transducer 1'); subplot(8,1,4); % plot(Time1,data1(1:(LMinute*60*10),4),'r-'); % title('The Raw Data: Channel 4: Force Transducer 2');

%%Calculation of the mean DC offset in every minute;

% figure(2); for i=1:LMinute OFST(i,1)=mean(data1(((i-1)*Fs*60+1):(i)*Fs*60,1)); % subplot(4,1,1); % plot([Time1((i-1)*Fs*60+1),Time1((i)*Fs*60)],[OFST(i,1), OFST(i,1)],'r-'); % title('DC offset for every minute'); % hold on; OFST(i,2)=mean(data1(((i-1)*Fs*60+1):(i)*Fs*60,2)); % subplot(4,1,2); % plot([Time1((i-1)*Fs*60+1),Time1((i)*Fs*60)],[OFST(i,2), OFST(i,2)],'r-'); % title('DC offset for every minute'); % hold on;

OFST(i,3)=mean(data1(((i-1)*Fs*60+1):(i)*Fs*60,3)); % subplot(4,1,3); % plot([Time1((i-1)*Fs*60+1),Time1((i)*Fs*60)],[OFST(i,3), OFST(i,3)],'r-'); % title('DC offset for every minute'); % hold on; OFST(i,4)=mean(data1(((i-1)*Fs*60+1):(i)*Fs*60,4)); % subplot(4,1,4); % plot([Time1((i-1)*Fs*60+1),Time1((i)*Fs*60)],[OFST(i,4), OFST(i,4)],'r-'); % title('DC offset for every minute'); % hold on; end

%%Eliminate the offset in every minute for i=1:LMinute data2(((i-1)*Fs*60+1):(i)*Fs*60,1)=data1(((i-1)*Fs*60+1):(i)*Fs*60,1)-OFST(i,1); data2(((i-1)*Fs*60+1):(i)*Fs*60,2)=data1(((i-1)*Fs*60+1):(i)*Fs*60,2)-OFST(i,2); data2(((i-1)*Fs*60+1):(i)*Fs*60,3)=data1(((i-1)*Fs*60+1):(i)*Fs*60,3)-OFST(i,3); data2(((i-1)*Fs*60+1):(i)*Fs*60,4)=data1(((i-1)*Fs*60+1):(i)*Fs*60,4)-OFST(i,4); end

for i=1:4 data2(:,i) = (data2(:,i)- min(data2(:,i)))/(max(data2(:,i)) - min(data2(:,i))); end % (Pow1 - min(Pow1))/(max(Pow1) - min(Pow1)) + 0.1; figure(1); subplot(8,1,2);hold on; % plot(Time1,data2(4800:(LMinute*60*10),1),'b-'); plot(Time(1:9000),data2(:,1),'b-');

% title('The Raw Data: Channel 1: EGG 1; Red: Before Offset; Blue: After Offset'); subplot(8,1,4);hold on; % plot(Time1,data2(4800:(LMinute*60*10),2),'b-'); plot(Time(1:9000),data2(:,2),'b-'); % title('The Raw Data: Channel 2: EGG 2; Red: Before Offset; Blue: After Offset'); subplot(8,1,1);hold on; % plot(Time1,data2(4800:(LMinute*60*10),3),'b-'); plot(Time(1:9000),data2(:,3),'b-'); % title('The Raw Data: Channel 3: Force Transducer 1; Red: Before Offset; Blue: After Offset'); subplot(8,1,3);hold on;

% plot(Time1,data2(4800:(LMinute*60*10),4),'b-'); plot(Time(1:9000),data2(:,4),'b-'); % title('The Raw Data: Channel 4: Force Transducer 2; Red: Before Offset; Blue: After Offset'); ymax12=max(max(max(data1(:,1)),max(data1(:,2))),max(max(data2(:,1)),max(data2(:,2)))); ymin12=min(min(min(data1(:,1)),min(data1(:,2))),min(min(data2(:,1)),min(data2(:,2)))); ymax34=max(max(max(data1(:,3)),max(data1(:,4))),max(max(data2(:,3)),max(data2(:,4)))); ymin34=min(min(min(data1(:,3)),min(data1(:,4))),min(min(data2(:,3)),min(data2(:,4)))); subplot(8,1,1);ylim([ymin12, ymax12]); subplot(8,1,2);ylim([ymin12, ymax12]); subplot(8,1,3);ylim([ymin34, ymax34]); subplot(8,1,4);ylim([ymin34, ymax34]);

data3(:,1)=power(data2(1:(LMinute*60*10),1),2); data3(:,2)=power(data2(1:(LMinute*60*10),2),2); data3(:,3)=power(data2(1:(LMinute*60*10),3),2); data3(:,4)=power(data2(1:(LMinute*60*10),4),2);

SMinute = 1; EMinute = 15; data4=data3(SMinute*60*Fs+1:EMinute*60*Fs,:);

Time1 = Time(1:9000); Time2=Time1(SMinute*60*Fs+1:EMinute*60*Fs);

%Normalize the plot figure(4); max1=max(data4(:,1)), min1=min(data4(:,1)); data4(:,1)=(data4(:,1)-min1)/(max1-min1); subplot(4,1,1);hold on; plot(Time2,data4(:,1),'b-'); max1=max(data4(:,2)), min1=min(data4(:,2)); data4(:,2)=(data4(:,2)-min1)/(max1-min1); subplot(4,1,2);hold on; plot(Time2,data4(:,2),'b-'); max1=max(data4(:,3)), min1=min(data4(:,3)); data4(:,3)=(data4(:,3)-min1)/(max1-min1); subplot(4,1,3);hold on;

plot(Time2,data4(:,3),'b-'); max1=max(data4(:,4)), min1=min(data4(:,4)); data4(:,4)=(data4(:,4)-min1)/(max1-min1); subplot(4,1,4);hold on; plot(Time2,data4(:,4),'b-');

Pow1 = [2.16 1.77 2.24 2.27 2.28 2.23 2.32 2.25 2.29 2.25 2.09 2.27 2.24 2.26 2.34]; Pow2 = [2.26 1.86 2.21 1.93 2.25 2.29 2.35 2.37 2.40 2.38 2.43 2.38 2.40 2.28 2.24]; Pow3 = [2.21 2.05 2.25 2.27 2.3 2.29 2.2 2.37 2.35 2.35 1.99 2.23 2.13 2.21 2.23]; Pow4 = [2.26 1.9 1.9 1.84 1.81 2.16 2.26 2.34 2.4 2.41 2.4 2.39 2.38 2.36 2.36];

% Pow1 = (Pow1 - min(Pow1) + 0.5) / (max(Pow1) - min(Pow1));

Pow1 = (Pow1 - min(Pow1))/(max(Pow1) - min(Pow1)) + 0.1;

Pow2 = (Pow2 - min(Pow2))/(max(Pow2) - min(Pow2)) + 0.1;

Pow3 = (Pow3 - min(Pow3))/(max(Pow3) - min(Pow3)) + 0.1;

Pow4 = (Pow4 - min(Pow4))/(max(Pow4) - min(Pow4)) + 0.1; %%Calculate Power Plot figure(1);hold on; for i=SMinute:(EMinute-1) POWV(i-SMinute+1,1)=sum(data4(((i-SMinute)*Fs*60+1):(i-SMinute+1)*Fs*60,1))/7; subplot(8,1,6);

x = 30:60:870; bar(x,Pow1,1.0,'r');

% plot([Time1((i-SMinute)*Fs*60+1),Time1((i-SMinute+1)*Fs*60)],[Pow1(i-SMinute+1), Pow1(i-SMinute+1)],'r-'); % title('Power Plot for EGG Channel 1'); hold on;

POWV(i-SMinute+1,2)=sum(data4(((i-SMinute)*Fs*60+1):(i-SMinute+1)*Fs*60,2))/7; subplot(8,1,8); bar(x,Pow2,1.0,'r'); % plot([Time1((i-SMinute)*Fs*60+1),Time1((i-SMinute+1)*Fs*60)],[Pow2(i-SMinute+1), Pow2(i-SMinute+1)],'r-'); % title('Power Plot for EGG Channel 2'); hold on;

POWV(i-SMinute+1,3)=sum(data4(((i-SMinute)*Fs*60+1):(i-SMinute+1)*Fs*60,3)) / 18; subplot(8,1,5);

bar(x,Pow3,1.0,'r'); % plot([Time1((i-SMinute)*Fs*60+1),Time1((i-SMinute+1)*Fs*60)],[Pow3(i-SMinute+1), Pow3(i-SMinute+1)],'r-'); % title('Power Plot for Force Transducer 1'); hold on;

POWV(i-SMinute+1,4)=sum(data4(((i-SMinute)*Fs*60+1):(i-SMinute+1)*Fs*60,4))/18; subplot(8,1,7);

bar(x,Pow4,1.0,'r'); % plot([Time1((i-SMinute)*Fs*60+1),Time1((i-SMinute+1)*Fs*60)],[Pow4(i-SMinute+1), Pow4(i-SMinute+1)],'r-'); % title('Power Plot for Force Transducer 2'); hold on;

end

ymax12=max(max(max(Pow1(:)),max(Pow1(:))),max(max(Pow2(:)),max(Pow2(:)))) + 0.4; ymin12=min(min(min(Pow1(:)),min(Pow2(:))),min(min(Pow2(:)),min(Pow2(:)))) - 0.4; ymax34=max(max(max(Pow3(:)),max(Pow3(:))),max(max(Pow4(:)),max(Pow4(:)))) + 0.4; ymin34=min(min(min(Pow3(:)),min(Pow3(:))),min(min(Pow4(:)),min(Pow4(:)))) - 0.1; ymin12 = -.1; ymin34 = -.1; ymax12 = 1.2; ymax34 = 1.2;

subplot(8,1,5);ylim([ymin12, ymax12]); xlim([0,900]); set(gca,'XTick',[ 0 100 200 300 400 500 600 700 800 900]) subplot(8,1,6);ylim([ymin12, ymax12]); xlim([0,900]); set(gca,'XTick',[ 0 100 200 300 400 500 600 700 800 900]) subplot(8,1,7);ylim([ymin34, ymax34]); xlim([0,900]); set(gca,'XTick',[ 0 100 200 300 400 500 600 700 800 900]) subplot(8,1,8);ylim([ymin34, ymax34]); xlim([0,900]);

set(gca,'XTick',[ 0 100 200 300 400 500 600 700 800 900])

[RHO13,PVAL13] = corr(POWV(:,1),POWV(:,3)); [RHO24,PVAL24] = corr(POWV(:,2),POWV(:,4)); [RHO14,PVAL14] = corr(POWV(:,1),POWV(:,4)); [RHO23,PVAL23] = corr(POWV(:,2),POWV(:,3)); [RHO34,PVAL34] = corr(POWV(:,3),POWV(:,4)); [RHO12,PVAL12] = corr(POWV(:,1),POWV(:,2));

% % string_time=strcat('Starting Second:',num2str(SMinute),' Ending Second:', num2str(EMinute)); % string_corr=strcat('RHO12:',num2str(RHO12),' RHO13:',num2str(RHO13),' RHO14:',num2str(RHO14),' RHO23:',num2str(RHO23),' RHO24:',num2str(RHO24),' RHO34:',num2str(RHO34)); % xlabel({string_time;string_corr}); jFrame = get(handle(gcf),'JavaFrame'); jFrame.setMaximized(true);

APPENDIX B: PCB design of TIIM receiver circuit