Function of the posterior in : In vivo laryngeal model

HONG-SHIK CHOI, MD, GERALD S. BERKE, MD, MING YE, MD, and JODY KREIMAN, PhD, Los Angeles, California

The function of the posterior cricoarytenoid (PCA) muscle In phonation has not been well documented. To date, several electromyographlc studies have suggested that the PCA muscle Is not simply an abductor of the vocal folds, but also functions In phonation. This study used an In vivo canine laryngeal model to study the function of the PCA muscle. SUbglottic pressure and electroglottographlc, photoglottographlc, and acoustic waveforms were gathered from fiVe adult mongrel dogs under varying conditions of stimulation. Subglottic pressure. fundamental frequency, sound Intensity, and vocal efficiency decreased with Increasing stimulation of the posterior branch of the recurrent laryngeal nerve. These results suggest that the PCA muscle not only acts to brace the against the anterior pull of the adductor and cricothyroid muscles, but also functions Inhlbltorlly In phonation by controlling the phonatory glottal width. (OTOLARYNGOL HEAD SURG 1993;109: 1043-51.)

The important physiologicfunctions of the larynx­ during phonation in some clinical cases. Kotby and protection of the lower airway, phonation, and res­ Haugen? also observed increased activity in the 1 piration - are all mediated by the laryngeal mus­ PCA muscle during phonation and postulated that cles. Intrinsic laryngeal muscles are classified into the muscle is not simply an abductor of the vocal three groups: the tensors, which regulate the length cord. and tension of the vocal folds; the adductors, which Gay et al." observed increased activity in the PCA close the ; and the abductor, which opens the muscle during phonation in chest at high glottis. The posterior cricoarytenoid (PCA) muscle pitches and concluded that the muscle plays a role is the sole abductor of the glottis and has generally as a tensor of the . Hirano" presented been considered a muscle of respiration.P EMG data from the PCA, the cricothyroid (CT), and Several authors have described the function of the the vocalis muscles during singing of ascending and PCA muscle in phonation.v" Faaborg-Anderson' descending scales in modal register by untrained found that electromyographic (EMG) activity in the male subjects. The EMG sample showed that the PCA muscle decreased during sustained phonation, PCA muscle was generally inactive during phona­ but started to increase a few milliseconds before the tion. However, some activity was noted at high end of audible voice. Brewer and Dana' recorded pitches, where the CT muscle was markedly acti­ increased activity from the PCA muscle when ex­ vated. With the use of a pharyngeal surface elec­ perimental animals attempted to whine. Ded06 and trode, Fujita et aLII also noted an increase in PCA Yang? reported increased activity in the PCA muscle muscle activityat the high end of the pitch range. Mu and Yang" observed EMG activity of the PCA muscle during phonation in 11 of 12 dogs. They From the Laryngeal PhysiologyLaboratory, Division of Head and concluded that the PCA muscle has a phonatory Neck Surgery, UCLA School of Medicine. function and that it may play an important role in Presented at the Annual Meeting of the American Academy of Otolaryngology-Head and Neck Surgery, Washington, D.C., precise glottis control. Sept. 13-17, 1992. The present study used an in vivo canine model of Received for publication Oct. 16,1992; revision received Feb. 5, phonation to study the role of the PCA muscle in 1993;.accepted March 5, 1993. phonation. This model has been described in detail Reprint requests: Gerald S. Berke, MD, Division of Head and elsewhere. 13.14 Phonation was initiated by stimulating Neck Surgery, UCLA School of Medicine, 10833 Le Conte Ave., Los Angeles, CA 90024. the anterior branches of the recurrent laryngeal Copyright © 1993by the American Academy of Otolaryngology­ nerve (RLN). The peAbranch of the RLN was then Head and Neck Surgery Foundation, Inc. stimulated to investigate the effect of the PCA 0194-5998/93/$1.00 + .10 23/10/46950 muscle on laryngeal vibration. 1043 Otolaryngology ­ Head and Neck SUrgery 1044 CHOI el 01. December 1993

METHODS The oral intubating tube was removed. With the use In Vivo Preparation of a small button and a 2-0 silk suture, the Five mongrel dogs were anesthetized with an was suspended for better visualization of the larynx. intramuscular injection of 3 ml acepromazine male­ An additional proximal tracheotomy was performed ate as a premedication, followed by intravenous through which a cuffed tracheotomy tube was sodium pentobarbital (Nembutal) titrated to loss of placed, with its tip resting 10 em below the glottis. the corneal reflex.Each animal wasplaced supine on The cuff of the superiorly directed tube was inflated an operating table and direct laryngoscopywas per­ to just seal the . Room air was bubbled

formed to confirm normal laryngeal anatomy. through 5 em H20 at 37° C for warming and hu­ A 7 mm oral endotracheal tube was inserted and midification and was then passed through the ceph­ connected to a respirator. After the animal was alad tracheotomy tube. shaved, prepared, and draped, a vertical midline All animal protocols were approved by the UCLA incision was made, and the strap muscles and ster­ Animal Care and Use Committee and were per­ nocleidomastoid muscles were retracted laterally to formed in compliance with the local, state, and expose the larynx and trachea. The external branch federal regulations for the humane use ofanimals in ofthe (SLN) was isolated at research. its entrance into the cr muscle. After twitch of the cr muscle was confirmed by nerve stimulation, Glottography. Pressure. and Intensity speciallydesigned rubber electrodes (custom-made, Measurements monopolar, flexible, conductive neoprene with sili­ Electroglottography electrodes (Synchrovoice, cone, coated with insulative silicone KE45W) were Briarcliff Manor, N.Y.) were placed in direct contact applied around the nerve (Fig. 1,B). The RLN was with the itself. The reference electrode was isolated at the tracheoesophageal groove and was sutured to the inside of the skin flap. confirmed with electrical stimulation. The inferior A photosensor (Centronics OSD 50-2, Mountain­ constrictor muscle was cut at the lateral margin of side, N.J.) was placed on the trachea approximately the cartilage. The was 3 em below the larynx.A halogen flashlight provided cut. The RLN was dissected superiorly to identify supraglottic illumination for photoglottography. the anterior and posterior branches and Galen's A catheter-tipped pressure transducer (Millar anastomosis after the posterior larynx was rotated. model no. SPC 330, Houston, Texas) was inserted The three branches (anterior, posterior, and Galen's through the upper tracheostoma and rested 2 em anastomosis) were confirmed with electrical stimu­ below the glottis. The transducer was calibrated lation. The anterior branch, which contracts the against a manometer from 0 to 100 mm Hg pressure adductor muscles, was cut just distal to the PCA before insertion. branch. A rubber electrode was attached to the Phonatory intensity was measured with a linear distal stump of the anterior branch proximal to the scale sound level meter (Quest Electronics model interarytenoid branching. Galen's loop was cut after no. 208L,Oconomowoc, Wis.),which was positioned confirmation. Another electrode was applied to the 30 em from the canine larynx. Photoglottography, trunk of the RLN for stimulation of the posterior electroglottography, subglottic pressure, and acous­ branch (Fig. 1, A). tic signals were low-pass filtered at 3 kHz and dig­ This procedure was then repeated on the opposite itized at 20 kHz with a 12-bit analog-to-distal con­ side of the animal. Electrodes attached to both sides version board. The signals were verified on an os­ of the SLN were connected to a Grass model 54H cilloscope (Tektronix 5116, Beaverton, Ore.) before stimulator (Grass Instruments, Quincy, Mass.). digitization, Files were stored on disks, and 0.5 to Electrodes attached to the anterior branches and 2.4-second samples of stable phonation were ex­ trunks of the RLN on both sides were connected to cerpted for analysis. A multipurpose computer pro­ separate channels of a second nerve stimulator (cus­ gram (CSpeech, version 3.1, P. Milenkovic, Madi­ tom-made, two-channel, constant-voltage, direct­ son, Wis.) was used to analyze subglottic pressure, current stimulator). A ground electrode was in­ glottography, and acoustic signals (Fig. 2). serted into the subcutaneous tissue of the neck flap. were stimulated by an 80 Hz pulse with 1.5 Vldeostroboscopy ms duration. Stimulation intensity varied from 0 to Stroboscopic imagesof vocal fold movements dur­ 2 volts, as will be described. ing phonation were recorded with the use of a Karl A distal tracheotomy was performed and an en­ Storz laryngostrobe unit (model 8000). A Storz zero­ dotracheal tube was connected to the respirator. degree telescope was connected to the stroboscope VTolaryngology - Head and Neck Surgery CHOt et 01. 1045 Volume 109 Number 6

A

TA BRANCH

LCA BRANCH

\\...._- - - RLN

Fig. 1. A, Posterior view of canine larynx. Anterior branch of RLN was cut just distal to peA branch (large arrow). 8, Photograph of canine larynx. Rubber electrodes applied to anterior branch (arrow) and trunk (open arrow) of RLN and external branch of SLN (arrowhead). TA, Thyroarytenoid muscle; LeA, lateral cricoarytenoid muscle; IA. interarytenoid muscle. Otolaryngology ­ Head and Neck Surgery December 1993 1046 CHOI et at

OireclCaFfJUterized Di9ilolizOlion Flow MeIer

Humidi'ie.!YamlAir

VWrlilolor

Fig. 2. SChematic drawing of experimental set-up. EGG. Electroglottography: PGG. photo­ glottography.

by means of fluid-filled cables. The image from the OQ. Separate regressions were undertaken for each telescope was recorded with a Storz CCD (charge­ dependent measure. coupled device) video camera (model 9000, Storz Static study. Stimulation to the posterior branch of Instruments, Culver City,Calif.) and a Sony U-matic the RLN was set at zero, low (0.3 volt), medium (0.5 videocassette recorder (VO-5800, Sony,Park Ridge, volt), and high (0.7 volt). Nerve stimulation re­ N.J.). mained constant at each level to induce a steady, stable voice. Trials were run with SLN stimulation Experimental Design set at a constant 1 volt and with no SLN stimulation, Dynamic study. Air flow remained constant at 388 as already described. Two to three trials were ob­ ml/sec throughout the experiment. Phonation was tained at each level of posterior branch stimulation. initially induced by stimulation of the anterior Trials were separated by at least 3 to 5 minutes to branch of the RLN. Initial stimulation to the ante­ reduce effects of fatigue. rior branch was set between 0.5 and 1 volt as nec­ Subglottic pressure, FO, intensity, 00, and vocal essary to produce as loud a voice as possible. Stim­ efficiency were analyzed for each level of posterior ulation to. the posterior branch of the RLN was branch stimulation. OQ and vocal efficiency were increased dynamically from 0 to 2 volts over a 3-sec­ calculated as described previously.F" ond trial. A second trial was then performed using Average values of subglottic pressure, FO, and OQ the same procedure, but with the addition of stim­ were calculated from 10consecutive cycles selected ulation to the SLN at a constant 1 volt. at random from a stable section of phonation from Data were evaluated at 300 ms intervals. Funda­ each trial (Fig. 3). Separate three-way analyses of mental frequency (FO), subglottic pressure, and variance were then calculated with these average open quotient (00) were averaged across 10 con­ values used as dependent measures. Independent secutive cycles for each 300msec interval. Multiple variables were subject (five dogs), PCA muscle stim­ regression was used to examine the effects of PCA ulation level (none, low, medium, high), and SLN muscle stimulation on FO, subglottic pressure, and stimulation level (absent or present). otolaryngology - Head and Neek Surgery Volume 109 Numb9f 6 CHOI et al. 1047

Scpeen Files Edit Analysis Recopd Play Quit I CH 3 8.218 VOLTS Final = 2836.358 Length = 61,358 Freq 1

Chl •••••_ ••••• ~l

Ch 3

Ch4

Fig. 3. One representative waveform from dynamic study: channel 1. subglottic pressure; channel 2, electroglottography; channel 3. photoglottography; and channel 4, acoustic waveform.

RESULTS 40] = 0.96, p > 0.01; 00: F[l, 32] = 49.08, Dynamic StUdy P < 0.01). Significant interactions between PCA Multiple regressions showed that FO, subglottic muscle and SLN stimulation also occurred for Fa, pressure, and OQ all declined significantly with PCA pressure, and 00. Because order of SLN stimula­ muscle stimulation level (FO: t = - 3.06, p < 0.01; tion conditions was randomized across animals, F[2, 67] = 4.85, P < 0.01; subglottic pressure: these interactions may reflect fatigue effects in some t = -2.90, p < 0.01, F[2, 67] = 9.49, p < 0.01; animals. Interactions may also be a result of ana­ OQ: t = -3.21, p < 0.01, F[2, 53] = 11.01, p < tomic differences among animals, which are inher­ 0.01; Figs. 4, 5, and 6). SLN stimulation condition ent in in vivo research. also significantly affected pressure and OQ, but not FO, independent of the effects of PCA muscle stim­ DISCUSSION ulation (Fa: t = 0.60, p > 0.01; pressure: t = The larynx acts as a variable resistor that regulates -3.25, p < 0.01; OQ: t = -3.43, p < 0.01). Pres­ airflow in and out of the lungs. Most variations in sure and OQ bot~ declined significantly in the pres­ laryngeal resistance occur at the level of the vocal ence of SLN stimulation, across PCA muscle stim­ folds: contraction of the PCA muscles abducts the ulation conditions. folds and lowers resistance, whereas contraction of the thyroarytenoid muscle and other vocal fold ad­ Static study ductors narrows the glottic slit and raises resis­ PCA muscle stimulation level had a significant ranee.':" effect on each dependent measure (Fa: F[3, 40] = Because abduction of the vocal fold results from 1774.86, P < 0.01; pressure: F[3, 40] = 2545.52, either electrical stimulation or physiologic induction P < 0.01; intensity: F[3, 40] = 246.03, p < 0.01; (e.g., inspiration), the PCA muscle was long consid­ OQ: F[3, 32] = 78.71, p < 0.01; Figs. 7, 8, and 9). ered to be simply an inspiratory or abductor muscle. Scheffe post-hoc comparisons showed that, across During quiet breathing in human beings, inspiration SLN conditions, each level of PCA muscle stimula­ is powered by the diaphragm and other muscles of tion differed significantly from all the others, for all the ventilatory bellows, and the PCA muscles con­ the dependent measures. Vocal efficiency (calculat­ tract to reduce entrance flow resistance. During ed as the ratio of acoustic power to subglottal power) expiration, flow is determined by the recoil pressure also decreased with increasing PCA muscle stimu­ of the , and there is no evidence of lation (Figs. 7 and 8). SLN stimulation condition PCA muscle contraction. However, in cases of in­ significantly affected FO, pressure, and OQ, but not creased ventilation as a result of hypercapnia or intensity (FO: F[l, 40] = 94.83, P < 0.01; pressure: exercise, a widely patent glottic airway is associated F[1, 40] = 1333.20, p < 0.01; intensity: F[l, with increased expiratory PCA muscle activity.":" OTolaryngology­ Head and Neck Surgery December 1993 1041 CHOI et al.

b a (sec) 0.0 0.3 0.8 0.9 1.2 1.5 1.8 2.1 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 (sec) 300 300

dogl 200 200 :~ - dog2 Hz Hz ---a- dog3 dog4 ~ ~ 100 ~ dogS ~ 100 ~ ~~ - 0 0 0 > 2V 0 > 2V

Fig. 4. Fundamental frequency tracing In dynamic stUdy with Increasing PCAmuscle stimulation: a, SLN off; b, SLN on.

b (sec) a 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 0.0 0.3 0.8 0.9 1.2 1.5 1.8 2.1 (sec) 200 200

150 150 dogl 0 0 ('II ('II - dog2 -tr- dog3 100 = - ua =a 100 U dog4 ;~ ~ dogS :: - 50 50 :~ - ~ ; ~ ::: 0 0 0 > 2V 0 :> 2 v Fig. 5. Subglottic pressure tracing In dynamic study with Increasing PCA muscle stimulation: a, SLN off; b, SLN on.

b a 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 (s e c) 1.8 2.1 0.0 0.3 0.6 0.9 1.2 1.5 (sec) 30 30

25 25 dog2 - dog3 "!o "!o --tr- dog4 20 20 -- dogS -- 15 15

10 10 0 :> 2 V 0 :> 2 V

Fig. 6. Open quotient tracing In dynamic study with Increasing PCAmuscle stimulation: a, SLN off; b, SLN on.

Recent advances in EMG recordings and process­ been reported that the intrinsic laryngeal muscles ing techniques have revealed several important participate actively in making glottal adduction-ab­ functions of the PCA muscle besides inspiratory duction adjustments during laryngeal articulatory contraction,":" Hirose" has reported extensively on gestures. In particular, the PCA muscle apparently the function of the PCA muscle in . He found plays a subtle role in glottal opening gestures, pro­ a marked elevation of PCA muscle activity for the vided there is simultaneous suppression of the ad­ production of voiceless and partial PCA ductor muscles. In other words, the principal mech­ muscle activation for voiced fricatives. It has also anism underlying adduction-abduction is reciprocal Otolal)'ngology - Head and Neck SurgeI)' CHOI et 01. 1049 Volume 109 Number 6

300 ~------...,

200 • Frequency(Hz)

Pressure(CmH20)

Intensity(dB)

~ Vocal elliciency(OE-05) 100

o o . Low Med High PCA stim u la t ion

Fig . 7. Static analysis ollundamentallrequency. subglottic p ressure, intensity, and vocal efficiency: SLN off .

300 ~------,

200 • Frequency(Hz) Pressure(CmH20)

Intensily (dB) 100 ~ Vocal elliciency(° E-05)

o o Low Med High P CA stimu la tion with SLN

Fig. 8. Static analysis of fundamental frequency, subglottic pressure, Intensity, and vocal efficiency: SlN on.

activation of the adductor and abductor groups of mucosa and that of the muscles are decreased and the larynx. Among the adductors, the interarytenoid passively stiffened." Thus the combined effect of muscle most clearly demonstrates a reciprocity with PCA muscle stimulation together with strong vocal the PCA muscle in terms of the voicing features in fold adduction on the FO was of interest. However, different ." lack of an in vivo animal model has made it difficult Hirano" has reported that when the PCA muscle to objectively measure the effects of the PCA muscle is stimulated electrically, the vocal processes are on vibration. In this study, frequency dropped as the abducted and elevated. The vocal fold becomes intensity of PCA muscle stimulation increased. markedly elongated and thin and the edge of the These decreases in frequency were apparently vocal fold rounded; the cross-sectional area of the caused by an increase in the posterior glottic gap Otolaryngology ­ Head and Neck Surgery 1050 CHOI et 01. December 1993

30..,....-....------,

20

% • noSLN

.. with SLN

10

o o Low Mad High peA stimulation

Fig. 9. Open quotient analysis in static study.

produced by PCA muscle stimulation. In compari­ mined by the pitch and phonetic nature of the son, pitch fell with increasing vibratory gap in a sounds being uttered.v" four-mass model," and pitch also decreased in an in During whisper, PCA muscle activity is increased vivo canine model of simulated RLN paralysis." relative to normal phonation, because glottal abduc­ In this study, subglottic pressure, vocal intensity, tion is increased.":" Whispered speech is generally and efficiency also decreased with increasing PCA less intense than normal voice, because of the effects muscle stimulation, in both the dynamic and static of prephonatory glottal width on the vocal inten­ studies (Figs. 5, 7, and 8). We hypothesize that the sity." When the prephonatory glottal gap is wide decrease in intensity is a result of the decrease in intensity is decreased, which is consistent with the subglottic pressure, which in turn is caused by the findings presented here. glottal gap produced by PCA muscle stimulation. Several studies have reported increases in PCA Thus the primary action of PCA muscle contraction muscle activity at higher pitch levels.":":" Research, seems to be (1) decreasing frequency and intensity ers have assumed this activity served either to brace of phonation and (2) reduction of glottal resistance the arytenoid against the strong anterior through decreases in subglottic pressure, although pull of " or to oppose increased there is some evidence of increased vocal fold ten­ adductor activation, thus increasing vocal fold stiff­ sion with PCA muscle stimulation. ness." Our data suggest that the PCA muscle not Although the PCA muscle may not playa primary only braces the arytenoids against anterior pull, but role in phonation, it appears important for fine also serves as an antagonist by regulating phonatory control of subglottic pressure, frequency, and inten­ glottal width. sity. Several authors have shown that motor unit Recently, Diamond et a1. 31 suggested that the PCA activity increases briefly but substantially in all vocal muscle has three bellies, each belly being composed fold adductor muscles just before each phonemic of different muscle fiber types and each having a utterance." These increases are accompanied by an different function. Further studies combining EMG equally brief decrease in the activity of the abductor and in vivo animal models may further elucidate the PCA muscles, with an interval that varies from 50 to functions of the PCA muscle in phonation. 25 27 500 msec. • Throughout an utterance, the vocal fold adductors show irregularly augmented motor CONCLUSION unit activity, as do the previously inhibited PCA An in vivo canine model was used to stUdy the muscles, the degree of augmentation being deter- function of the PCA muscle during phonation. In- Otolaryngology - Head and Neck Surgery Volume 109 Number 6 CHOI et 01. 1051

creases in PCA stimulation produced a decrease in vivo canine model. Ann Otol Rhinal Laryngol 1991;100: fundamental frequency, subglottic pressure, inten­ 280-7. 15. Proctor OF. Physiologyof the upper airway. In: Handbook of sity, and vocal efficiency. This inhibitory function of physiology. Vol. 1. Washington, D.C.: American Physiologi­ the PCA muscle in phonation suggests that it not cal Society, 1964:309-45. only has a bracing action against the anterior pull of 16. Barlet OJ. Effects of hypercapnia and hypoxia on laryngeal the adductor and cricothyroid muscles, but also resistance to airflow. Respir Physiol 1979;37:293-302. increases phonatory glottal width. 17. Barlet OJ, Remmers JE, Gautier H. Laryngeal regulation of respiratory airflow. Respir Physiol 1973;18:194-204. We wish to thank Bruce Gerratt, PhD, for his assistance 18. Murakami Y, Kirchner JA. Respiratory movements of the in statistical analysis. vocal cords: an electromyographic study in the cat. Laryngo­ scope 1972;82:454-67. REFERENCES 19. Hirose H. Posterior cricoarytenoid as a speech muscle. Ann 1. Sasaki cr. Physiology of the larynx. In: English GM, ed. Otol Rhinol Laryngol 1976;85:334-42. Otolaryngology. Vol. 3. Philadelphia: JB Lippincott, 1984:1­ 20. Hirose H. Laryngeal articulatory adjustments in terms of 26. EMG. In: Hirano M, Kirchner JA, Bless OM, eds, 2. Hast MH. The respiratory muscle of the larynx. Ann Otol Neurolaryngology: recent advances. Boston: College-Hill, Rhinol LaryngoI1967;76:489-97. 1987:200-8. 3. Brancatisano TP,Dodd OS, Engel LA. Respiratory activityof 21. Hirano M. The laryngeal muscles in singing. In: Hirano M, posterior crico- and vocal cords in humans. Kirchner JA, Bless DM, eds. Neurolaryngology: recent ad­ J Appl Physiol 1984;57:1143-9. vances. Boston: College-Hill, 1987:209-30. 4. Faaborg-Anderson K. Electromyographic investigations of 22. Smith ME, Berke GS. The effects of phonosurgery on laryn­ intrinsic laryngeal muscles in humans. Acta Physiol Scand geal vibration-part 1: theoretic considerations. OTOLARYN­ SuppI1957;41:1-148. GOl HEAD NECK SURG 1990;103:380-90. 5. Brewer OW, Dana ST. Investigations in laryngeal physiol­ 23. Trapp TK. Berke GS, Bell TS, Hanson DG, Ward PH. Effect ogy- the canine larynx: part 2. Ann Otol Rhinol Laryngol of vocal fold augmentation on laryngeal vibration in simu­ 1963;72:1060-75. lated recurrent laryngeal nerve paralysis: a study of Teflon 6. Dedo HH. The paralyzed larynx: an electromyographic study and phonogel. Ann 0101 Rhinol Laryngol 1989;98:220-7. in dogs and humans. Laryngoscope 1970;80:1455-517. 24. Slavit DH, McCaffrey TV, Yanagi E. Effect of superior 7. Kotby MN, Haugen LK. Critical evaluation of the action of laryngeal nerve on vocal fold function: an in vivo canine the posterior crico-arytenoid muscle, utilizing direct EMG­ model. OTOlARYNGOl HEAD NECK SURG 1991;105:857-63. study. Acta Otolaryngol (Stockh) 1970;70:260-8. 25. Buchthal F, Faaborg-Andersen KL. Electromyography of 8. Gay T, Hirose H, Strome M, Sawashima M. Electromyogra­ laryngeal and respiratory muscles. Ann Otol Rhinol Laryngol phy of the intrinsic laryngeal muscles during phonation. Ann 1964;73:118-23. Otol Rhinol Laryngol 1972;81:401-10. 26. Wyke BD. Recent advances in the neurology of phonation: 9. Yang S. Electromyography of the physiological functions of phonatory reflex mechanisms in the larynx. Br J Disord the intrinsic laryngeal muscles. Chin J Otorhinolaryngol Commun 1967;2:2-14. 1984;19:69-72. 27. Hirano M, Vennard W.Ohala J. Regulation of register, pitch 10. Hirano M. Vocal mechanisms in singing: laryngological and and intensity of voice: an electromyographic investigation of phoniatric aspects. J Voice 1988;2:51-69. intrinsic laryngeal muscles. Folia Phoniatr 1970;22:1-20. 11. Fujita M, Ludlow CL, Woodson GE, Nauton RF. A new 28. Kotby MN. Haugen LK.The mechanics of laryngeal function. surface electrode for recording from the posterior crico­ Acta Otolaryngol (Stockh) 1970;70:203-11. arytenoid muscle. Laryngoscope 1989;99:316-20. 29. Sawashima M, Hirose H. Laryngeal gesture in speech pro­ 12. Mu L, Yang S. The role of posterior crico-arytenoid muscle duction. In: MacNeilage PF, ed. The production of speech. in phonation: an electromyographic investigation in dogs. New York: Springer-Verlag, 1983:11-38. Laryngoscope 1991;101:849-54. 30. Titze IR. Regulation of vocal power and efficiency by sub­ 13. Berke GS, Moore OM, Hanson DG, Hantke DR, Gerratt glottal pressure and glottal width. In: Fujimura 0, ed. Vocal BR, Burstein F. Laryngeal modeling: theoretical, in vitro, in fold physiology. New York: Raven Press, 1988:227-38. vivo. Laryngoscope 1987;97:871-81. 31. Diamond AJ, Goldhaber N, Wu BL, Sanders I. The intra­ 14. Green DC, Berke GS, Ward PH. Vocal fold medialization by muscular nerve supply of the posterior cricoarytenoid muscle surgical augmentation versus arytenoid adduction in the in of the dog. Laryngoscope 1992;102:272-6.