A Case Report in Changes in Phonatory Physiology

Following Voice Therapy: Application

of High-Speed Imaging
*Rita R. Patel, Jack Pickering, *Joseph Stemple, and Kevin D. Donohue, *zLexington, Kentucky and yNew York, New York

Summary: Objective. To clinically evaluate changes in vocal fold vibration and voice production caused by voice
therapy in hoarseness resulting from contact granuloma.
Design. Single-subject before-after prospective study using multiple measures of vocal function. A 6-week program of
vocal function exercises (VFEs) was conducted using multiple assessments of vocal function to identify and measure the
changes pre- and posttreatment, in a 51-year-old male with unilateral contact granuloma. Multiple outcome measures
were recorded. High-speed digital imaging (HSDI) measures of voice onset time (milliseconds), open quotient, speed
quotient, maximum amplitude, peak closing velocity, peak-to-average opening velocity, and peak-to-average closing ve-
locity were derived from motion data. Acoustic measures of maximum phonation duration (seconds), noise-to-harmonic
ratio, average fundamental frequency (hertz), the lowest fundamental frequency (hertz), and the highest fundamental
frequency (hertz); aerodynamic measures of expiratory volume (milliliter) and mean expiratory airflow (liter/second);
stroboscopic measures of glottal closure and phase closure; and perceptual assessment of voice quality (total score) using
the Consensus Auditory-Perceptual Evaluation of Voice were obtained.
Results. Stroboscopic, acoustic, aerodynamic, and audioperceptual measures were minimally informative related to
pre- and posttreatment vocal function in a patient with contact granuloma. HSDI measures provided multiple physiologic
and kinematic measures demonstrating pre- and posttreatment efficiency of vocal function, including vibratory motion,
closure, and impact stress.
Conclusion. The results have implications for the use of high-speed imaging to identify and measure change in pho-
natory physiology in patients with contact granuloma. Changes in phonatory physiology support the use of voice therapy
techniques, such as VFEs that facilitate a semioccluded vocal tract for treatment of contact granuloma.
Key Words: High-speed imagingVoice therapyVocal physiologyContact granuloma.

INTRODUCTION process contact during phonation, voice therapy is often recom-

A vocal fold granuloma or a contact granuloma is a benign, re-
active/reparative lesion along the posterior one-third of the vo-
cal fold that presents as granulation tissue or fibrosis overlying
intact or ulcerated squamous epithelium along the vocal pro-
cess1 of the arytenoid cartilage. A contact granuloma is thought
to result from trauma due to acid reflux, intubation, vocal hyper-
function and / or large peak interarytenoid pressures.2 Insuffi-
cient glottal closure owing to vocal fold stiffness and vocal
fold paresis is also known to result in contact granuloma.3,4
More males appear to develop vocal fold granuloma than fe-
males.5 Presenting symptoms from contact granuloma can
range from no vocal symptoms; to lump in the throat sensation,
pain, and shortness of breath; to severe hoarseness. Owing to
the impact stress resulting from repetitive trauma from vocal
Accepted for publication January 10, 2012.
Funding: none.
From the *Division of Communication Sciences and Disorders, Department of Rehabil-
itation Sciences, University of Kentucky, Lexington, Kentucky; yDepartment of Commu-
nication Sciences and Disorders, The College of Saint Rose, New York, New York; and the
zDepartment of Electrical and Computer Engineering, University of Kentucky, Lexington,
Kentucky.
Address correspondence and reprint requests to Rita R. Patel, Division of Communica-
tion Sciences and Disorders, Department of Rehabilitation Sciences, University of
Kentucky, 900 South Limestone, 120 D, Charles T. Wethington Building, Lexington,
KY 40536-0200. E-mail: rita.patel@uky.edu
Journal of Voice, Vol. 26, No. 6, pp. 734-741
0892-1997/$36.00
2012 The Voice Foundation
doi:10.1016/j.jvoice.2012.01.001
mended as a primary mode of treatment6 coupled with manage-
ment of laryngopharyngeal reflux.1,4,79 Botox injection and/or
surgical excision is occasionally recommended to prevent re-
currence.4,8 A comprehensive review of vocal fold granuloma
is provided by Hoffman et al.10
There is little information in the literature regarding the vi-
bratory characteristics resulting from contact granuloma.
Through high-speed film recordings, von Leden and Moore11
first identified large movements of the arytenoids with in-
creased closed time resulting in impact stress at the level of
the vocal process. An anterior glottal gap2,12 identified in cases
with vocal fold granuloma is also thought to result in high con-
tact stress at the level of the vocal process.
Laryngeal visualization of vibratory patterns is fundamental
for appropriate assessment and treatment of most voice disor-
ders; however, empirical investigations of vibratory changes
as a response to vocal fold granuloma resulting in impact stress
are not known. Vocal symptoms of low pitch voice, monotone,
vocal fry, hyperfunction,1315 and vocal effort16 identified per-
ceptually and confirmed by acoustic and aerodynamic assess-
ments are thought to result from increased impact stress
resulting in contact granuloma. Although valuable, acoustic
and aerodynamic measurements provide indirect estimates of
vocal function. How these translate to laryngeal biomechanics
and vibratory features on laryngoscopy is not known. Hillman
et al13 using acoustic and aerodynamic assessments of voice,
Rita R. Patel, et al Changes in Phonatory Physiology Following Voice Therapy
revealed abnormally high levels of AC flow, peak flow, and
maximum flow declination rate in two cases with contact gran-
uloma, suggesting high vocal fold closure velocities, high colli-
sion forces, abnormally large amplitude, with incomplete
closure resulting in large open quotient. Empirical physiologi-
cal investigations directly quantifying these spatiotemporal fea-
tures of high closing velocities, large amplitude, and large open
quotient from vocal fold motion are not known. Clinically, de-
tailed accurate assessments of spatiotemporal vibratory features
for the entire vocal fold image can now be obtained with the use
of high-speed digital imaging (HSDI) as it captures actual
cycle-to-cycle vibratory motion.17,18
The purpose of this study was to comprehensively investigate
changes in vocal physiology before and after voice therapy, in
a patient with contact granuloma using multidimensional as-
sessments of HSDI, stroboscopy, acoustic analysis, aerodynam-
ics, and perceptual assessments of voice quality.
METHODS
Participant
A 51-year-old college professor and speech-language patholo-
gist with a history of laryngopharyngeal reflux and a left vocal
process granuloma was recruited for the study at the University
of Kentucky Vocal Physiology and Imaging Laboratory, after
signing an institutional review board-approved informed con-
sent. Diagnosis of contact granuloma was made 4 months before
the initiation of the study on otolaryngological examination.
The patients chief complaint included vocal fatigue, hoarseness
after teaching a long class, and occasional lump in the throat
sensation in the morning. The patient was also diagnosed with
laryngopharyngeal reflux and right vocal fold sulcus vocalis
along the anterior one-third margin of the right vocal fold. Since
the diagnosis, the patient was taking a proton-pump inhibitor
once a day and followed antireflux dietary and behavioral pre-
cautions, with minimal change in voice quality or lesion size,
as monitored via monthly stroboscopic examination of the vocal
folds. Owing to persistent symptoms of hoarseness and patient
requirements of heavy professional voice use, voice therapy
was recommended.
Voice therapy program
A well-established voice therapy program with proven out-
comes19 known as vocal function exercises (VFEs) was imple-
mented for 6 weeks. The VFE20 voice therapy approach is a type
of physiologic voice therapy designed to improve: (1) the bal-
ance among the subsystems for voice production, (2) laryngeal
muscle strength, (3) voice control and stamina, and (4) supra-
glottic modification of the laryngeal tone. The exercises consist
of the following 4 steps: (1) warm-up, (2) stretchinggliding
from the lowest note to the highest note, (3) contracting
gliding from the highest note to the lowest note, and (4) adduc-
tory powersustaining five musical notes (C-D-E-F-G) for as
long as possible on vowel o while pursing the lips to partially
occlude the vocal tract. The patient performed the four steps of
the VFE, twice daily, once in the morning and once in the even-
ing for 6 weeks.
735
Study design
In this prospective case study, multiple measures were used to
assess changes of vocal function consequent to a voice therapy
approach. HSDI, stroboscopy, and perceptual assessment of
voice quality were obtained pre- and post-VFE. In addition to
the above, acoustic and aerodynamic measurements were ob-
tained pre- and post-VFE and at 1 week interval. Percentage
change in vocal function was calculated for the measurements
derived from stroboscopy, HSDI, acoustics, aerodynamics,
and perceptual assessments before and after VFE, only for mea-
surements that were not within the normal limits.
Data collection and measurements
Acoustic analysis. Weekly and pre- and posttherapy acous-
tic measurements were computed for sustained phonation on
the vowel /a/ with the KayPentax (Montvale, NJ) Computerized
Speech Laboratory (CSL), model no. 4500. Multidimensional
voice profile module of the KayPentax CSL system was used
for analysis. A constant mouth-to-microphone distance of
6 in. was maintained for the recordings. A SHURE PG48
(Niles, IL), cardoid dynamic microphone of professional qual-
ity was used to obtain the acoustic recordings. The following
parameters were computed for sustained phonation: maximum
phonation time (MPT, seconds), average fundamental fre-
quency (hertz), jitter (%), low fundamental frequency (hertz),
high fundamental frequency (hertz), shimmer (decibels), and
noise-to-harmonic ratio.
Aerodynamic analysis. Pre- and posttherapy aerodynamic
measurements of expiratory volumes (milliliter) and mean ex-
piratory airflow (liters per second) were computed using the
KayPentax Phonatory Aerodynamic System no. 6600.
High-speed digital imaging. For the HSDI recordings,
a KayPentax high-speed system model 9710 was used. Images
were recorded at 4000 frames/s for a maximum duration of
4.094 seconds with a spatial resolution of 512 3 256 pixels.
A 300 W xenon light source was used. Recordings were made
on sustained phonation of the vowel /i/ at typical pitch and loud-
ness and on the laryngeal diadochokinetic task of repeating /he
he he he/ four times with short breaths between syllable. Three
trials of all the tasks were recorded for pretreatment and four tri-
als recorded for posttreatment. Pre- and posttherapy spatiotem-
poral features of voice onset time, open quotient (duration of
the cycle during which the vocal folds remain open divided
by the duration of the entire cycle),21 maximum amplitude,
speed quotient (duration of lateral movement divided by the du-
ration of medial movement), relative peak closing velocity,
peak-to-average opening velocity, and peak-to-average closing
velocity were derived from the vibratory motion using custom-
developed image processing software using Matlab (Math-
Works, Natick, MA).22 Voice onset time was calculated in
milliseconds as the duration from initial vocal fold edge motion
after complete abduction to the first vocal fold contact which is
followed by steady-state cyclic oscillations, typically preceded
by prephonatory irregular motion. The voice onset time was
calculated using glottal cycle montage analysis combined
with digital kymography of the task /he he he/.
736 Journal of Voice, Vol. 26, No. 6, 2012

TABLE 1.
Mean (Standard Deviations) of the Spatiotemporal Measurements Derived from High-Speed Digital Imaging Before and
After 6 weeks of Voice Therapy
Spatiotemporal Vibratory Measures (Units) Prevoice Therapy (Right/Left) Postvoice Therapy (Right/Left)
Voice onset time (ms) 140 (0.05) 77 (0.06)
Maximum amplitude (pixels) 0.078 (0.006) 0.100 (0.006)
Open quotient NA 0.6739 (0.0089)
Speed quotient 0.92 (0.08)/0.58 (0.02) 0.44 (0.04)/0.48 (0.03)
Peak closing velocity (glottal lengths/s) 18.5 (1.1)/18.0 (1.3) 27.1 (3.4)/24.0 (1.1)
Peak-to-average opening velocity 3.44 (0.04)/2.73 (0.12) 1.37 (0.04)/1.58 (0.17)
Peak-to-average closing velocity 2.11 (0.06)/1.62 (0.05) 1.69 (0.19)/1.45 (0.11)
Pitch and intensity were kept constant for pre- and postvoice therapy recordings.

The spatiotemporal features of maximum amplitude, speed ration and phase closure was performed using standard strobo-
quotient, relative peak closing velocity, peak-to-average opening scopic imaging at 30 frames/s on sustained phonation of the
velocity, and peak-to-average closing velocity were extracted vowel /i/ at comfortable pitch and loudness. Three blinded
from the high-speed video sequence using semiautomatic edge speech-language pathologists performed pre- and posttherapy
detection methods to track vocal fold edge displacements from ratings of stroboscopic imaging using a consensus based
the medial line in pixels. The analysis was performed for the approach.
midmembranous section of the vocal fold for steady-state pho- Audio-perceptual assessment. Three blinded speech-
nation at normal pitch and loudness. Displacements for the right language pathologists performed pre- and posttherapy voice
and left vocal folds were extracted from the detected edges, assessments of overall voice quality (total score) using the Con-
which were verified by direct human observation (edges overlaid sensus Auditory-Perceptual Evaluation of Voice (CAPE-V).23
on video images). The glottal length in pixels was estimated
through manually marking frames in the video sequence. The
glottal lengths were used to normalize displacement values to RESULTS
reduce the variability inherent in the pixel dimensions, which Pre- and posttreatment changes in vocal physiology in a single
was used for some of the features (Table 1). The spatiotemporal subject with unilateral contact granuloma was investigated with
features were computed for every cycle and the median value measures from multiple voice assessment domains, including
over all cycles was used to characterize motion for the given acoustics, aerodynamics, HSDI, laryngeal videostroboscopy,
steady-state phonation. Table 1 represents the mean and the and audioperceptual assessment. The participant was enrolled
standard deviation values for features computed over the inde- in a 6-week voice therapy program (VFE) in which progress
pendent recordings. Maximum amplitude is the maximum dis- was determined by improvement in MPT from baseline to
placement of the vocal fold during the opening phase, relative completion of the exercises at week 6. Table 2 demonstrates
to the glottal length, and is computed for each cycle as the improvement in mean MPT of 35.21 seconds at baseline to
sum of the right and left vocal fold peak displacement divided 45.1 seconds at week 6.
by the glottal length in pixels. The peak closing velocity is de-
fined as the minimum displacement gradient value during the Acoustic analysis
closing phase of the glottal cycle. Simple gradient differences Acoustic measures of average fundamental frequency, jitter%,
of normalized adjacent displacement points in the closing phase low and high extent of fundamental frequency, and noise-to-
were used to compute the peak closing velocity. The velocity is harmonic ratio were within normal limits before and after
in units of glottal lengths per second because the displacements
were normalized by the glottal lengths. The peak-to-average
opening velocity characterizes velocity evenness or change dur- TABLE 2.
ing the opening phase. A stronger peaking (higher value) reflects Weekly Averages of Maximum Phonation Times for the
a greater acceleration at shorter intervals during the motion. Vocal Function Exercises in Seconds
Peak-to-average opening velocity is the maximum opening
Musical
velocity divided by the average of all velocities in the open Note Week 1 Week 2 Week 3 Week 4 Week 5 Week 6
phase. Because this feature is a ratio of velocities, it is therefore
unitless. The peak-to-average closing velocity is analogous to C3 35.25 37.57 44.93 46.61 44.25 42.43
D3 35.64 38.07 47.21 49.11 46.04 45.14
the peak-to-average opening velocity but represents the change
E3 35.14 38.50 48.75 50.29 46.29 45.57
in velocity during the closing phase. F3 35.43 38.29 48.36 49.21 46.61 45.89
Stroboscopic imaging. Stroboscopic assessment was per- G3 34.61 36.64 45.18 47.89 45.11 46.64
formed using KayPentax Digital Stroboscopy system RLS Average 35.21 37.81 46.89 48.62 45.66 45.13
9100b. Visual perceptual assessment of glottal closure configu-
Rita R. Patel, et al Changes in Phonatory Physiology Following Voice Therapy 737

TABLE 3.
Acoustic Data for Sustained Vowel /a/ Weekly Before and After (Week 6) VFE
Acoustic Measures Pre-VFE Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Norms, Mean (SD)
Average F0 (Hz) 134.1 134.6 136.7 131.0 137.1 151.5 132.4 145.22 (23.41)
Jitter (%) 0.363 0.240 0.180 0.150 0.190 0.228 0.252 <0.59 (0.54)
Low F0 (Hz) 70.0 57.4 58.8 63.6 65.4 72.0 67.6 77482*
High F0 (Hz) 468.5 596.0 580.3 596.0 558.2 580.9 496.3 77482*
Relative intensity level (dB) 60.55 70.62 69.79 67.12 62.66 74.77 79.80 <60110*
Shimmer (dB) 0.20 0.25 0.15 0.16 0.26 0.16 0.20 <0.22 (0.085)
NHR 0.086 0.141 0.152 0.145 0.164 0.143 0.109 <0.12 (0.014)
Source: The norms are taken from the KayPentax Computerized Speech Labs database within the Multidimensional Voice Program.
Abbreviations: VFE, vocal function exercise; SD, standard deviation; F0, fundamental frequency; NHR, noise-to-harmonic ratio.
* The norms marked are taken from Stemple JC, Glaze L, Klaben B. Clinical Voice Pathology, Theory and Management. 4th ed. San Diego, CA: Plural Publishing;
2010:178.

therapy (Table 3). The greatest pre- posttest acoustic change open quotient for the posttreatment was 0.674. The mean speed
was demonstrated in mean intensity level. A 26.7% (60.55 quotient was also reduced with the right fold showing greater
79.80 dB) increase in intensity level was noted after VFE. reduction (0.900.48) than the left fold (0.580.53). The
mean peak closing velocity in glottal lengths per second was in-
Aerodynamic analysis creased for both folds (18.5 to 27.1 right; 18.0 to 24.0
Aerodynamic measures were taken pre- and posttreatment left) as the vocal folds were moving greater distances over
while keeping both frequency and intensity constant (Table 4). shorter period of time. The mean peak-to-average velocity
The participant demonstrated a 4.07% increase in respiratory (this is unitless) during both the opening and closing phases de-
volume (190 mL) and a 54.16% decrease in mean expiratory creased after treatment for both the vocal folds, with the right
airflow following 6 weeks of VFE. decreasing more than the left vocal fold (Table 1). For all the
spatiotemporal measures, the standard deviations were low
HSDI results (Table 1), being one order of magnitude less than the pre-
Greatest percentage change in pre- and posttherapy measure- and posttreatment changes for all except the speed quotient of
ments on high-speed imaging was observed for voice onset the left fold. This suggests that the variability among the
time, maximum amplitude, speed quotient, open quotient, repeated measurements is smaller than the variability cased
peak closing velocity, peak-to-average opening velocity, and by therapy.
peak-to-average closing velocity (Table 1). Voice onset time
was reduced from 140 to 77 milliseconds. The mean maximum Laryngeal stroboscopy results
amplitude of vibration was increased from 7.8% to 10% of the Pretreatment, glottis closure was characterized by a small ante-
glottal length. The vocal folds remained open during the pre- rior glottal gap and a complete posterior closure. Posttherapy
treatment recording; hence, the calculations of open quotient demonstrated improved closure of the membranous vocal folds
are not applicable for the pretreatment recordings. The mean and a small posterior glottal gap. Figure 1 demonstrates both the
glottis closure configuration and the phase closure pattern of the
vocal folds pre- and postvoice therapy. The open and closed
TABLE 4.
phases of the vibratory cycle appeared to be near equal for
Aerodynamic Measurements Before and After 6 Weeks of both pre- and posttreatment.
Voice Therapy
Audio-perceptual results
Prevoice Postvoice
The overall voice quality scores of the CAPE-V were calculated
Aerodynamic Measures Therapy Therapy Norms
and averaged from pre- and posttreatment. The mean overall se-
Expiratory volume (mL) 5480 5670 >2000 verity rating on the CAPE-V was 18.33/100 mm pretherapy in-
Mean expiratory 0.12 0.065 0.080.2 dicative of a mildly deviant voice quality. Posttherapy, the mean
airflow (L/s)
overall severity rating on CAPE-V was 0/100 mm indicative of
Mean sound pressure 76.68 79.80
level (dB)
a normal voice quality.
Mean fundamental 171.73 173.85
frequency (Hz) DISCUSSION
Pitch was maintained constant for the pre- and postvoice therapy record- The goal of this study was to evaluate changes in voice pro-
ings as indicated by the value of mean fundamental frequency. duction and vocal physiology before and after voice therapy in
A 3 dB difference in intensity level was noted for pre- and postvoice
therapy.
hoarseness resulting from contact granuloma. Multiple voice
assessment domains of HSDI, laryngeal videostroboscopy,
738
FIGURE 1. Pre- and postvoice therapy glottal closure configuration
from stroboscopic assessment revealing small posterior phonatory gap
following voice therapy, suggestive of reduced impact stress along the
vocal process.
acoustics, aerodynamics, and audioperceptual parameters, were
used to compare the vocal function of an individual with unilat-
eral vocal fold granuloma, pre- and postvoice therapy. Most re-
vealing was the contribution of HSDI in expanding the authors
understanding of the clinical implications of vocal process gran-
uloma including impact stress and changes in vibratory dynam-
ics resulting from treatment. For example, multiple routine
acoustic measures failed to demonstrate abnormalities in voice
quality or vocal function pre- or posttreatment in spite of the par-
ticipants complaints of rough voice and vocal fatigue. Indeed,
jitter and shimmer measures and noise-to-harmonic ratios
were within normal limits throughout the testing as were funda-
mental frequency and frequency range. The only pre- posttest
differences in the acoustic measures were seen in relative inten-
sity level and an increase in the high extent of fundamental
FIGURE 2. Digital kymographs showing improved voice onset time after 6 weeks of successful voice therapy. The number 1 indicates the first mo-
tion of the vocal fold after complete abduction and 2 indicates the first vocal fold contact of steady-state cyclic oscillations after the prephonatory state.
Journal of Voice, Vol. 26, No. 6, 2012
frequency. These results bring into question the efficiency of
routine acoustic measures for demonstrating characteristics of
vocal function with this patient with contact granuloma.
The selected aerodynamic measures, expiratory volume, and
mean expiratory airflow showed a minor increase in volume and
a decrease in airflow rate. It can be speculated that the VFE
trained the subject to breathe to his maximum capacity, thus in-
creasing volume. The decrease in airflow rate was consistent
with improved midfold glottal contact during adduction, which
was demonstrated with only high-speed imaging posttreatment.
The most illuminating assessment tool for demonstrating
pre- and posttreatment changes in vocal function appeared to
be HSDI. The measurements from HSDI clearly demonstrated
differences in spatiotemporal characteristics of not only the vo-
cal fold vibration in pre- and posttreatment, but also between
the left and right vocal folds. The posttreatment phonation of
this subject with unilateral (left) contact granuloma clearly
demonstrated a decreased onset time, which is illustrated in
the kymographic images taken from the high-speed vibratory
image (Figure 2). Typical values of voice onset time for young
adults (less than 78 years) was reported to be 77.5 millisec-
onds.24 After treatment, the participants voice onset time
reached within the normal range. Improved vocal flexibility
was clearly demonstrated by increased maximum amplitude
of vibration posttreatment. In addition, the open quotient and
speed quotient measures posttreatment demonstrate more bal-
ance in the timing and phases (open/closed) of the vibratory cy-
cle (Figure 3). The open quotient after treatment was
approximately 0.67, which is closer to the normal of 0.64
0.88.21 Before treatment, the open quotient could not be com-
puted as the glottis remained open during the vibratory cycle.
The speed quotient, especially of the right vocal fold, was closer
to the value of 1, before treatment. This indicates that the open-
ing and closing parts of the glottal cycle for the right vocal fold
were of equal duration, when for normal motion the vocal folds
Rita R. Patel, et al Changes in Phonatory Physiology Following Voice Therapy 739

FIGURE 3. Pre- and postvoice therapy glottal cycle montage from high-speed digital imaging revealing small posterior phonatory gap and
improved midmembranous vocal fold contact following voice therapy, suggestive of reduced impact stress along the vocal process.

open faster than they close, resulting in a speed quotient ap-
proximating the reported ranges from 0.50 to 0.78.21
Finally, and perhaps most revealing in this case of unilateral
contact granuloma, HSDI revealed that the peak-to-average ve-
locity function of both vocal folds was decreased from pre-
posttreatment thus reducing the impact stress of the vibrating
vocal folds. Before treatment, there was no closure of the ante-
rior to the midmembranous fold during the glottal cycle, with
the posterior portion of the vocal fold remaining closed for
the entire cycle (Figure 4). This finding of incomplete closure
along the membranous part of the vocal fold provides support
to the theory of a presence of a large anterior glottal gap2,12 re-
sulting in increased contact stress along the vocal process. The
more pronounced velocity peaks observed during both the
opening and closing phases before treatment (Figure 5) suggest
an uneven force on the moving vocal folds. These sharp veloc-
ity changes could be a result of an imbalance in the muscle
control of the membranous part potentially owing to high
impact stress along the vocal process. Reduced mean values
of the peak-to-average opening velocity for posttreatment
(Figure 6) especially during the closing phase suggest a consis-
tent force on the vocal fold with more balanced muscle control
for vocalization.
Spatiotemporal features of impact stress from high-speed im-
aging in this case of granuloma were increased speed quotient,
increased voice onset time, and increased peak-to-average
opening/closing velocity. Maximum midmembranous ampli-
tude and peak closing velocity was smaller before treatment,
which is contrary to the inference of vocal function from previ-
ous study.13 One explanation for this could be that the right vo-
cal fold appeared stiff with the presence of potential sulcus
vocalis, resulting in reduced vibratory amplitude and reduced
peak closing velocity at the midmembranous portion. Moreover
the vocal folds remained open during the entire cycle along the
740 Journal of Voice, Vol. 26, No. 6, 2012

FIGURE 4. A. Indicates the pretherapy size of the vocal fold granuloma. B. and C. Indicates glottal closure pattern before and after voice therapy
in a case of contact granuloma from high-speed digital imaging. Note an incomplete glottal closure along the entire membranous part of the vocal
fold with complete closure of the posterior glottis before therapy.

mid membranous and anterior part of the vocal folds, which
would suggest that the vocal folds are traveling a shorter dis-
tance and hence the reduced peak closing velocity measure-
ment. It would be interesting to calculate the peak closing
velocity along the posterior part of the vocal fold, at the site
of the granuloma. The vocal folds however remained closed
along the posterior part for the entire cycle during the pretreat-
ment phase and hence this measurement could not be calcu-
lated. After voice therapy, a consistent small posterior
phonatory gap was maintained at the posterior part of the vocal
folds, which along with the improved closure of the membra-
nous vocal fold, could be inferred as reduced impact stress
along the posterior margin. The other possibility could be that
the impact stress resulting in vocal fold granuloma is different
from the impact stress resulting in vocal fold nodules, where
one would expect large vibratory amplitude at the midmembra-
nous section of the vocal fold. It may be that velocity functions
ranging from peak-to-average are more representative of im-
pact stress for in vivo measurements of vocal function. How-
ever, large studies would need to be conducted to determine
the spatiotemporal measurements of impact stress and their nor-
mative ranges in controlled phonation. The values of reduced
speed quotient, voice onset time, and peak-to-average open-
ing/closing velocity are unique to this study. Although further
large-scale study is clearly warranted, the data here suggest
the usefulness of HSDI for clinical investigation of vocal phys-
iology owing to the impact stress in patients with vocal fold
granuloma.

A6 A6
Percent glottal length

4
Percent glottal length

4
2
2
0

0 2

2 4

6
4 100 102 104 106 108 110 112 114 116 118 120
100 102 104 106 108 110 112 114 116 118 120
Milliseconds
Milliseconds

B
B 6
6
Percent glottal length

4
4
2
2
0
0
2
2
4 Velocity/10
4 Velocity/10 % Amplitude
% Amplitude 6
6 100 102 104 106 108 110 112 114 116 118 120
100 102 104 106 108 110 112 114 116 118 120 Milliseconds

FIGURE 5. Pretreatment midglottal vocal fold edge waveforms for
three cycles. A. Edge detection points right (3 marker) and left
(B marker) with denoised waveform as solid line in units of percent
glottal length. Note no contact of vocal folds during the cycles (B). Ve-
locity estimate for right fold scaled down by a factor of 10 to superim-
pose on the amplitude displacement waveform. Note multiple peaking
of velocities during the closing and opening phases of a glottal cycle.
FIGURE 6. Posttreatment midglottal vocal fold edge waveforms for
three cycles. A. Edge detection points right (3 marker) and left (B
marker) with denoised waveform as solid line in units of percent glottal
length. Note complete contact of vocal folds during the cycles (B). Ve-
locity estimate for right fold scaled down by a factor of 10 to superim-
pose on the amplitude displacement waveform. Note the smooth
velocities during the closing and opening phases of a glottal cycle.
Rita R. Patel, et al Changes in Phonatory Physiology Following Voice Therapy
Beyond the computational measures of the HSDI were the
visual observations afforded by this assessment tool. It has pre-
viously been reported that voice therapy approaches, such as
VFEs and resonant voice therapy promote a glottal closure pat-
tern (barely adducted/abducted arytenoids) that enhances effi-
cient vocal fold vibration through the use of a semioccluded
posturing of the vocal tract.25 Pre- and posttreatment high-
speed images clearly demonstrate this glottal configuration.
Figure 4A shows an incomplete glottal closure along the ante-
rior and midmembranous part of the vocal fold with complete
closure of the posterior glottis prevoice therapy. Figure 4B
demonstrates a different closure after therapy with improve-
ment in glottis closure along the midmembranous vocal folds
and a small posterior glottal gap suggesting reduced posterior
compression. A montage of vibratory cycles demonstrating
pre- and posttreatment vibratory patterns and glottal closure
configuration are demonstrated in Figure 3. The results of this
single-subject study suggest the utility of HSDI in demonstrat-
ing changes in vibratory physiology pre- and postvoice therapy.
The results further suggest that large-scale studies investigating
changes in glottal closure patterns as a function of exercises that
use the semioccluded vocal tract are warranted.
CONCLUSION
In this prospective study using multiple measures to clinically
evaluate vocal function, it was determined that visual percep-
tual and quantitative measurements of vocal function provided
by HSDI were best in identifying features of vocal function
compared with the stroboscopic, acoustic, aerodynamic, and
audio-perceptual assessments in a case with contact granuloma.
Stroboscopy was useful in providing insight regarding glottic
closure patterns, which are often implicated in voice disorders
and phonatory changes resulting from voice therapy. HSDI
was the only assessment technique that provided multiple phys-
iologic and kinematic measures clearly demonstrating (1) im-
pact stress and (2) pre and post-treatment changes in vocal
function in this patient with unilateral contact granuloma. In
a single subject, this study demonstrates the use of HSDI for
clinical assessment of impact stress and for measuring out-
comes after intervention. Future large-scale studies using re-
sults from HSDI as primary outcome measures are warranted.
REFERENCES
1. Devaney KO, Rinaldo A, Ferlito A. Vocal process granuloma of the
larynxrecognition, differential diagnosis and treatment. Oral Oncol.
2005;41:666669.
2. Hess MM, Verdolini K, Bierhals W, Mansmann U, Gross M. Endolaryngeal
contact pressure. J Voice. 1998;12:5067.
741
3. Eller R, Marks L, Hawkshaw M, Sataloff R. Vocal process granuloma. Ear
Nose Throat J. 2007;86:198.
4. Carroll TJ, Gartner-Schmidt J, Statham MM, Rosen CA. Vocal process
granuloma and glottal insufficiency: an overlooked etiology? Laryngo-
scope. 2009;120:114120.
5. Kiese-Himmel C, Kruse E. Sociodemographic variables of a German sam-
ple of patients with contact granuloma. J Voice. 1995;9:449452.
6. Scheid S, Anderson T, Sataloff R. Nonoperative treatment of laryngeal
granuloma. Ear Nose Throat J. 2003;82:244245.
7. Damrose EJ, Damrose JF. Botulinum toxin as adjunctive therapy in refrac-
tory laryngeal granuloma. J Laryngol Otol. 2008;122:824828.
8. Leonard R, Kendall K. Effects of voice therapy on vocal process granu-
loma. Am J Otolaryngol. 2005;26:101107.
9. Bloch CS, Gould WJ, Hirano M. Effect of voice therapy on contact granu-
loma of the vocal fold. Ann Otol Rhinol Laryngol. 1981;90:4852.
10. Hoffman H, Overholt E, Karnell M, McCulloch T. Vocal process granu-
loma. Head Neck. 2001;23:10611074.
11. von Leden H, Moore P. Contact ulcer of the larynx. Arch Otolaryngol.
1960;72:746752.
12. Kleinsasser O. Microlaryngoscopic and histologic appearances of polyps,
nodules, cysts, Reinkes edema, and granulomas of the vocal cords. In:
Kirchner J, ed. Vocal Fold Histopathology: A Symposium. San Diego,
CA: College Hill Press; 1986:5155.
13. Hillman R, Holmber E, Perkell J, Walsh M, Vaughan C. Objective assess-
ment of vocal hyperfunction: an experimental framework and initial results.
J Speech Hearing Res. 1989;32:373392.
14. Boseley ME, Myers KV. Conservative treatment of an obstructing vocal
fold granuloma. Ear Nose Throat J. 2003;82:550.
15. Ylitalo R, Hammarberg B. Voice characteristics, effects of voice therapy,
and long-term follow-up of contact granuloma patients. J Voice. 2000;14:
557566.
16. Verdolini K, Hoffman H, McCoy S. Nonspecific laryngeal granuloma:
a case study of a professional singer. J Voice. 1994;8:352358.
17. Deliyski D, Hillman R. State if the art laryngeal imaging: research and clin-
ical implications. Curr Opin Otolaryngol Head Neck Surg. 2010;18:
147152.
18. Patel R, Dailey S, Bless D. Comparison of high-speed digital imaging with
stroboscopy for laryngeal imaging of glottal disorders. Ann Otol Rhinol
Laryngol. 2008;117:413424.
19. Stemple J, Lee L, DaMico B, Pickup B. Efficacy of vocal function exercises
as a method of improving voice production. J Voice. 1994;8:271278.
20. Stemple JC, Glaze L, Klaben BG. Clinical Voice Pathology: Theory and
Management. San Diego, CA: Singular; 2000.
21. Timcke R, Von Leden H, Moore P. Laryngeal vibrations: measurements of
the glottic wave. I. The normal vibratory cycle. AMA Arch Otolaryngology.
1958;68:119.
22. Unnikrishnan H, Donohue D, Patel R. Analysis of high-speed digital
phonoscopy pediatric images. Proc. SPIE 2012;8207:82071Q.
23. Kempster G, Gerratt B, Verdolini Abbott K, Barkmeier-Kraemer J,
Hillman R. Consensus auditory-perceptual evaluation of voice: develop-
ment of a standardized clinical protocol. American Journal of Speech-
Language Pathology. 2009;18:124132.
24. Kunduk M, Yan Y, McWhorter AJ, Bless D. Investigation of voice initiation
and voice offset characteristics with high-speed digital imaging. Logoped
Phoniatr Vocol. 2006;31:139144.
25. Berry DA, Verdolini K, Montequin D, Hess MM, Chan R, Titze I. A quan-
titative output-cost ratio in voice production. J Speech Lang Hear Res.
2001;44:2937.