77 / 412
Croake et al
after the University of Iowa head and neck protocol for
LEMG diagnostics.
21
Laryngeal electromyography was
performed on the right thyroarytenoid muscle of 7 partici-
pants with normal vocal function on 3 separate occasions
with the application of an additional behavioral control for
vocal intensity. We hypothesized that time of testing and
vocal intensity would significantly alter quantitative mea-
sures of thyroarytenoid EMG signals.
Materials and Methods
Seven participants, ages 18 to 40 years (3 male and 4
female), were recruited and volunteered for this study. Ten
participants were initially recruited based on power analysis
for the repeated measures design. Two participants did not
complete the entire study protocol, and 1 participant had
unusable data. All were nonsmokers, English speaking, and
free of laryngeal pathology and movement disorders as
judged by stroboscopic examination of the larynx. Exclusion
criteria were professional voice training, diagnosed bleed-
ing disorder, prior head/neck/spinal surgery, and/or intuba-
tion within the past 5 years. All procedures were approved
by the University of Kentucky Institutional Review Board,
and informed consent was obtained from all participants.
All assessments were performed at the University of
Kentucky Laryngeal & Speech Dynamics Laboratory.
Prior to data collection, all participants received a
laryngeal exam via videostroboscopy to ensure normal
vocal function and structure (Kay Elemetrics Rhino-
Laryngeal Stroboscope [Model RLS 9100 B], Kay
Elemetrics 70 degree rigid endoscope [Model SN 1541];
PENTAX Medical, Montvale, New Jersey, USA). This
examination was performed by a certified/licensed speech-
language pathologist and reviewed by a board certified
otolaryngologist.
After imaging was completed, electromyographic biopo-
tentials (μV) were acquired from the right TA muscle via a
25-mm, 30-gauge concentric bipolar needle electrode
(XLTEK 101468; Natus Medical, San Carlos, California,
USA). The raw EMG signal was routed to a biopotential
amplifier (Grass Model 15A54; Natus Neurology, Warwick,
Rhode Island, USA) and serially coupled to a 16-bit analog-
to-digital converter sampling at 10 kHz (PowerLab 1630;
ADInstruments, Inc, Colorado Springs, Colorado, USA).
Prior to digitization, analog signals were band-pass filtered
(30 Hz–3 kHz @ –3 dB), with an in-line notch filter applied
to reduce 60 Hz line contamination. The LEMG analog out-
put was paralleled to a stereo amplifier and played over
loud speakers to provide the otolaryngologist with auditory
feedback of muscle activity during needle insertion. Audio
(volts) and sound intensity levels (dB SPL) from the partici-
pant’s vocalizations were recorded by a lapel-style micro-
phone (Sony ECM44B; Sony Corporation, New York, New
York, USA) and a commercially available sound level meter
(REED ST-8850; REED Instruments, Sainte-Anne-De-
Bellvue, Quebec, Canada). Both signals were also digitized
by the A/D system (audio sampling rate = 2 kHz; sound
pressure level = 1 kHz). All signals were recorded in cali-
brated units using a proprietary 2-point interpolation
method found in our digitization software package
(LabChart 7; ADInstruments, Inc). Post-acquisition LEMG
signal processing was completed using custom-coded
LabChart routines.
Laryngeal electromyography testing was performed in a
custom-built Faraday booth to reduce electromagnetic field
effects. Participants were seated in an examination chair,
reclined to approximately 60 degrees, with their heads
comfortably supported by a neck pillow. No sedation or
anesthetic was used during the needle insertion and record-
ing procedure. A ground electrode was placed on the par-
ticipant’s neck, below the mastoid process. Thyroarytenoid
needle insertion by an otolaryngologist with 10 years of
experience performing clinical LEMG procedures was
accomplished using a para-medial approach percutane-
ously with the electrode directed in a superolateral direc-
tion through the cricothyroid ligament with the muscle
entered submucosally. Needle electrode placement was
confirmed using the following behavioral tasks: normal
rest breathing, phonation on sustained /i/, sniff, and sus-
tained phonation of falsetto /i/. Upon completion of the
study, the needle electrode was removed by the physician
and participants were monitored for 15 minutes after the
study in case of complications.
Digitized raw LEMG signals were full-wave rectified
and RMS signal amplitude values were calculated online
with the LabChart software package. Root mean square was
calculated as the square root of the mean of a series of
squared LEMG amplitude values. Root mean square ampli-
tudes were used in all analyses to determine if significant
variance in the LEMG signal existed as a function of vocal
intensity and time of data acquisition.
Experimental Protocol
Electromyography of the right TA was performed on each
participant on 3 different occasions with a minimum of 1
month between adjacent procedures (mean duration = 2.5
months) to allow for tissue healing. Each session took place
at approximately the same time of day and lasted no more
than 30 minutes (including videostroboscopy). After needle
electrode placement was confirmed within the TA, LEMG
signals were recorded under 2 task conditions. First, a con-
firmation condition was completed by recording baseline
LEMG signals while the participant was instructed to relax
and breathe normally, sustain the vowel /i/ at a comfortable
modal pitch, gently sniff through the nose, and sustain a
falsetto /i/. Second, a feedback condition was completed
with the participant producing a sustained /i/ at 2 different
57