.
.
.
.
.
.
.
.
This is a whole body pressure
plethysmograph. It's designed to contain one speaker.
And crucial in the design characteristics are that it holds
the person comfortably without too much extra space,
that the walls be completely rigid and non-yielding, and
that it be airtight. So we made ours out of
one and a half steel drums, and
it has the following design characteristics.
We added a lip at the bottom to increase the stability.
Here's the two parts of the
two different steel drums welded together. We also added a lip
at the top to mesh with the lid.
The lid is attached with quick release
levers, and the lid itself is
plexiglass so we can see the person, the person can see inside.
There is a face mask through which the subject breathes external air,
so that actually the ultimate seal between the body of the person and the outside
is around the nose and chin area
made by the flange of the face mask. So this shows how the plethysmograph
works. This is schematically the plethysmograph,
this inner ellipse is the speaker,
and of course the speaker is connected to the outside through the face mask.
This simulates a pressure transducer. So when the person
is at the end of an
expiration, most of the lung volume is at its minimum,
there is a low pressure developed in the air around the person, so the
pressure is low. When however the person is at the height of
breathing in, has taken the greatest inspiration, their total body
volume increases, particularly the chest, and that puts greater
pressure on the air around the outside, so a greater pressure is developed.
Now scoot down as far as you can.
Okay, hug your
knees. Okay, so now up we go.
Okay, you can stand up now.
Let's see.
Alright, turn around the other way.
Okay, now you can
turn around the other way.
Okay, comfortable? Alright, now we put the lid back on.
Maybe you want to duck now.
Okay.
Okay.
Now we tighten the lid.
And it's
very
important to have an airtight
seal, and if there's any leak detected, a little bit of Vaseline will help out.
Okay, now we
connect some air pressure tubes.
These are quick attaching
valves, which maintain complete airtight seal.
Are you comfortable?
It's going to be hot and muggy inside.
It will only last for a short time.
Say the first few words.
Again.
Can you try some Hindi?
Again.
Good. Okay, I'll now attach the pneumo attacker graph so we can measure airflow.
Alright.
Okay.
Okay.
Now we're going to try something different.
We're going to try to measure
sub-glottal pressure.
And this requires you to remove the face mask.
Okay.
You can just
hold on to it. And now hold your hand
over the hole. Okay, so you're now breathing
a closed system. So whatever air you take in under pressure leaves lesser
pressure on the outside. And as you breathe out, reducing the pressure
inside, you're augmenting the pressure in the air around you.
So now say, pin twice and so on.
Okay.
Let's let you have some fresh air.
So, feel better?
Yes. So here we have the
plathismograph signal at the top and the airflow
signal at the bottom. These are averages of ten tokens.
And the synchronization point for that
averages is the release of the D in this
sentence, you drew will near. Of course this is a
funny sentence designed just to have mostly sonorance, just has one
stop in it. This is the point where the pressure,
the lung volume changes from being increasing to decreasing, of course,
is the start of the expiration on which speech is superimposed.
Of course it generally shows a downward trend because speech is
made on an expiration. So the airflow signal
shows a decrease in airflow during the closure of the D, and then a sudden
increase upon its release. Corresponding with that we can see a slight
plateau, a slight leveling off of the lung volume decrement
at the moment of the D, otherwise it's fairly smooth.
At the end of speaking you usually have more air
left in your lungs than you need, you expire it and it comes out in high
volume and at that point the lungs start a kind of passive collapse
and then move on into the breathing in inspiration.
So this is the sentence you drew pill near, and the
interest here is of what happens during the aspirated stop.
Of course again the airflow shows a decrease during the closure,
and then a huge increase in airflow upon the release of this
stop since it's aspirated. And again during the closure there's a slight
plateau in lung volume decrement and then during the moment of
aspiration there's a steep decline. So I view this as a passive
collapse of the lungs during the
aspirated stop.
And here's when Priscilla was simulating some Hindi stops,
Ibhi, it shows much the same character for the lung volume
curve as it did for the voiceless aspirated stop,
a sharp decrease upon the moment of release, which is where the aspiration
occurs. In this case we've also taken the first derivative of this
curve, and that's what this is again averages.
And it turns out that this curve pretty much, if we may use
this curve from the voiceless stop,
that the airflow curve pretty much is a mirror image of
the differentiated lung volume curve.
Okay.
Okay.
Now if you want to have those newspapers here.
In our speech production laboratory here at the University at Buffalo,
we collect aerodynamic as well as respiratory kinematic data produced during
speech. Our primary goal is to describe and investigate children's speech
mechanisms with an emphasis on the laryngeal and respiratory systems.
In order to do this in a relatively non-invasive way, we use several
sensing devices that the children really tolerate quite well. The first device is called
a pneumotachograph mask and is used for sensing airflow.
This particular mask design can be described as being
circumferentially vented and was developed by Martin Rothenberg at Syracuse University.
As you can see, it's made from a standard respiratory face mask, which has
a series of holes punched around the circumference of the mask. The holes are
covered by a fine wire mesh, and that presents the slight resistance to the flow of
air through the mask. The resistance causes a pressure differential, which is
proportional to the airflow through the mask. Right here
is a pressure tap that's also a port to the air pressure transducer,
which is mounted right on the mask.
This particular transducer has a wide band frequency
response in that it picks up higher frequency flow components than most typical air pressure transducers.
It's because of this response that it produces a waveform that looks very much like a voicing
waveform. Notice that we have a second pressure tap,
which leads to a second transducer with a lower frequency response. It's from this transducer
that we generate the average airflow and the volume waveforms.
The third tap on the mask right here is connected to another pressure transducer
optimized for measuring intraoral air pressure. Inserted
inside the mask at this tap is also a small
hollow polyethylene tube that will be placed into the oral cavity and picks up
the intraoral air pressure values.
Now I'm going to demonstrate how we calibrate for airflow, air pressure, and air volume.
To calibrate for the airflow, we place the pneumatograph
on the mold. While using a device called a rotometer
to indicate the flow moving through the mask transducer system, we record this signal
and use it for later reference.
We calibrate for airflow for each mask size
that we use. So this is the medium size mask and this is the one that's used for adult females
and teenagers. We use the large size mask for adult males
and this is the small size mask that's used for the younger children.
When calibrating for intraoral air pressure, we use a device
that's referred to as a U-tube water manometer. By displacing
the water column and measuring the distance
that the column moves on each side, we can express the subsequent pressure
measurements in terms of centimeters of water. We record the calibration signal produced
and use it for later measurement.
Our next calibration procedure is for making the volume measurements. By not having our subjects breathe
into a respirometer to calibrate for volume, we don't fatigue them with an additional calibration maneuver.
This has been especially helpful when working with the children or with hard to test
patients. This time we connect the mask on the mold
to the respirometer and we repeatedly displace two liters of air
through the mask. We digitally integrate this cyclic flow signal
and then scale our low frequency flow channel.
This allows us to calculate the volume displacement by integrating the flow signal.
This brings me to the last piece of equipment that I'd like to demonstrate, the linearized
magnetometers. This is just a brief deviation from the aerodynamic techniques that
I've already presented. But since we use this quite typically with our
aerodynamic techniques, I wanted to present it at the same time. The magnetometers have been used for
quite a while and you can look at Hixson and Huyten-Hixson for further detail.
Linearized magnetometers operate in pairs. There are two pairs for sensing the chest wall movements,
one for the rib cage and the other for the abdomen. So that when attached to the front and back
of the abdomen, they will track the movement of that part of the chest wall.
Here I'm just going to demonstrate
the movement of the abdominal pair.
The emphasis on this
demonstration so far has been on speech aerodynamics and now I'm briefly going to
demonstrate just how we collect the simultaneous respiratory kinematic data from
unwilling but well-paid children. Here's a segment from a tape made of Nick, my son,
performing the paw paw paw task.
We've already attached the magnetometers to the chest wall. We've got one pair attached to the rib cage
and the other pair attached to the abdomen. Turn around, Nick,
so that we can look at that. Alright, now turn around this way.
What we're going to do next is have Nick sit in a chair and then we'll look at how
we pair the aerodynamic equipment with the respiratory kinematic equipment.
We're ready to put the mask on Nick.
Before we start, I'd like to point out that we've got this part of the mask which would hit the bridge of
the nose built up with a little bit of foam on the inside because very often we get a leak
in this area and we want to make sure there's not a leak during the task. Okay, Nick, I'm going to put this on your face.
I'm going to kind of smoosh it on hard and then I want you to put the tube in your mouth like you
would a straw, right behind your lips and behind your teeth and then you'll just talk around it.
And you're going to say that paw, paw, paw task for me, okay?
Okay. Ready, go.
Okay.
You're going to say seven of those. I want you to make them all connected and I want you to make a nice paw.
Not a baa, but a paw. Okay, ready, go.
Paw, paw, paw, paw, paw, paw.
Alright, at this point we would work with the subject. We would make sure that the
number of syllables were the correct number, that there was a tight seal around the edge of the mask.
Let me just take this off. That we had appropriate pressure waveforms
and that generally things are just the way we want them. We'd try to get three trials
for this particular task. Then if we wanted to do some inverse filtering, we might have Nick
phonate an ah or an aah and look at inverse filtering during a sustained vowel.
And all the while that we're collecting the data, as I said before, we'll check for
leaks and for the appropriateness of the task. And that's it.
Now let's take a look at some of the waveforms that were produced by that session.
The first waveform is the wideband airflow signal
which includes voicing. It starts off with rest breathing, indicated by
the slower alternations and then proceeds into the speech task, paw, paw, paw.
The second waveform is the low frequency airflow waveform, also called average
airflow. Next is the intraoral air pressure waveform, which demonstrates
the pressure variations as the subject is saying paw, paw, paw.
It's from these two waveforms, the average airflow waveform and the intraoral
air pressure waveform that we make estimates of laryngeal airway resistance, all of Smytherin and Hixson.
The fourth waveform represents the subject's lung volume,
which is the result of the integration of the average airflow.
The fifth and sixth waveform show the movement of the rib cage and the abdomen.
On each of these last three waveforms, the lung volume waveform, the rib cage and the
abdomen, we can look at where the speech utterance was initiated and where
the utterance was terminated. Finally, there's the voicing waveform,
which has been calibrated for sound pressure level and is used for SPL measures
during the speech tasks. Let's take a closer look at the top airflow waveform.
Here we have expanded the mid-portion of this vowel.
This signal came from the wideband pressure transducer and includes the
fundamental frequency as well as the formant frequency information.
The bottom waveform on the figure shows the oral airflow waveform
after it's been inverse filtered using C-Speech, a computer program developed by
Milankovic. It's now depicted as a glottal airflow waveform,
and it's from this waveform that we make a variety of measures, such as
approximations of how long the vocal folds are open relative to the duration of the cycle,
the speed at which the airflow shuts off, the constant flow that may be
present due to something like a posterior opening between the vocal folds during the closed part of the cycle.
Previous research in this area has been published by scientists such as
Holmberg, Hillman & Prokow, Goffin & Sundberg, and Tietze.
I'd like to finish by restating that the emphasis in our laboratory has been the simultaneous
measurement and interpretation of laryngeal and respiratory function.
We primarily work with children using aerodynamic and respiratory kinematic
techniques. There are, of course, a variety of ways that you could use these techniques clinically.
For example, if we wanted to help evaluate the speech of someone with a cleft palate,
we would enlist the aerodynamic techniques for looking at nasal flow and intraoral air
pressure. If we wanted to use the procedures with neurogenically impaired individuals,
we might use the respiratory kinematic measures to evaluate respiratory function,
as well as the aerodynamic techniques to help quantify velopharyngeal function.
We might also use a combination of measures to evaluate and help plan a voice
evaluation. We could study and quantify vocal intensity, fundamental
frequency, and glottal airflow measures. Lastly, the system
is flexible. A clinician can decide what to do based on each patient's needs.
An accurate
physical
description of
speech production requires precise measurements of moving
air and tissue.
This video demonstrates how these measurements are collected and how these
properties fit together in a mathematical model.
The elastic properties of laryngeal tissues are quantified by stretch-release
experiments.
Laryngeal tissue is mounted in a double-walled glass chamber containing a physiologic
solution as demonstrated by Dr. Yang Min.
Water is circulated between the double walls of the glass chamber to maintain
the tissue at body temperature. Perfusion provides proper oxygenation to the
tissue during the experiment. At the bottom, the tissue sample is secured
within the chamber. At the top, a servo-controlled cantilever arm provides
mechanical stimulus to the tissue. With this arm, specified
elongations or forces can be applied to the tissue in a stretch-release
pattern. The data are recorded automatically by computer, and the entire
cyclic stress-strain curve can be plotted immediately after the experiment.
These data provide the basic information for elastic properties of vocal fold
tissue. The experiments can be done on the cover of the vocal fold, as well as
on the ligament and muscle. Pressure flow events are quantified by conducting
experiments in a scaled-up model of the glottis. Once key measurements have been
made, the Navier-Stokes equation is solved to numerically supplement any information
that cannot be obtained empirically. A long fluid channel conditions the
air so that the flow in the subglottal region is laminar and can be analyzed in a
two-dimensional profile. Pressure levels are recorded in the subglottal, glottal,
and supraglottal regions using fine taps that are drilled into the static
vocal fold configuration. In addition, hot-wire anemometers are used to measure
the velocity profile of the air. Dr. Fahri Alapur takes measurements to quantify
airflow through the glottis and the pressure flow relationships. This
includes quantification of vortex shedding, turbulence, and instability in the
jet stream. Once the fundamental constitutive equations for tissue and air
movement have been established, it is important to quantify the geometric shape
of the vocal fold. First, Brad's story demonstrates how a reasonable
representation of the pre-phonatory glottis is made. Silicone is injected
into the glottal space of an excised larynx and allowed to harden, producing a
mold of the glottal airway. Molds are extracted and sliced into various layers
to get geometric measurements of the glottis in the superior-inferior
direction and the anterior-posterior direction. Using measurements gleaned from
the mold, the glottis is modeled with a mathematical formula. We have found it
appropriate to model the lateral surface of the vocal folds with a parabola and
the anterior-posterior variation with a straight line. Next, stroboscopy is used
to quantify the modes of vibration in the tissue, including amplitudes of
vibration as a function of subglottal pressure and fundamental frequency. A
large data set has been developed which shows how amplitude and frequency are
affected by vocal fold elongation and lung pressure. The modes of vibration are
simulated by computer. Three-dimensional graphics allow the model to be viewed
from different perspectives. A primary objective in physiologic modeling of
laryngeal function is to show how intensity, fundamental frequency, and
quality in phonation change with various muscle contractions. Dr. Harry Hoffman
places hooked wire electrodes into the laryngeal muscles with a hypodermic
needle after local anesthesia. The mic-to-mouth distance is measured so that
sound output levels between subjects can be compared. The subject then executes an
extensive set of vocal exercises. When measures of subglottal pressure are
needed simultaneously with cricothyroid and thyroarytenoid activity, a tube is
placed into the corner of the mouth to measure oral pressure in a bilabial
consonant-vowel sequence. In this manner, subglottal pressure can be estimated
from oral pressure in techniques that are described in literature. An important
graphic representation of the results of these experiments is the muscle
activation plot, which shows cricothyroid activity versus thyroarytenoid
activity with subglottal pressure as a parameter. Constant fundamental
frequency bands are drawn that indicate the combinations of various activities
that can produce the same fundamental frequency. The next step in the modeling
process is to compare simulated glottographic signals with those that can
be measured noninvasively. Specifically, mouth airflow and
electroglottographic signals are used to validate the model. Kenneth Tom
demonstrates how simultaneous measures of mouth flow and EGG are taken. The
signals are displayed in real time on an oscilloscope and later digitized for
examination and processing. A similar series of combined waveforms simulated
by the model is compared to measured waveforms. An optimization procedure is
used to match the two sets of waveforms, thus refining the parameters of the
model. A final step is needed to validate the model. Here, vibratory modes in
human subjects are contrasted with those produced by the model. The best results
from human subjects are obtained with a rigid fiber scope placed into the mouth
toward the back of the throat, as Eileen Finnegan demonstrates. It is possible to
make measurements of vibrational amplitude and, to a limited extent, mucosal
wave propagation and vertical phase differences in the movement. It is also
possible to see higher order modes that may occur due to peculiar excitations.
Wireframe diagrams and animations provide a means to study modes of the
model in detail.
Once a simulation model has been validated, we have a rich new medium for
experimentation. In fact, the model itself becomes a new laboratory. We're
currently using the model to study phonosurgical procedures, to study voice
disorders, to distinguish between artistic and normal vocal productions, and
to study control of pitch, loudness, register, and voice quality in speech.
The beauty of a simulation model is that you can't break it and it doesn't
deteriorate, and no human subject or animal is at risk in the exploration.
Okay, are you ready to start again today? Good. We're going to do the same thing
as last week. We're going to work on the phrase, how are you? We'll do it two
times. The first time we're going to emphasize are, how are you? The second
time we're going to emphasize you, how are you? Okay, good. Put that on again
today. Ready? Okay. How are you? How are you? Good.
This deaf college student is working on improving her voice using a device
called the electric autograph, or EGG. The EGG permits non-invasive examination
of time varying changes in vocal fold contact during voice production. This
easily obtained physiologically based measure can provide clinicians and
researchers with useful information about vocal fold morphology and
physiology that augment less direct acoustic measures or more invasive
techniques such as direct laryngoscopy. The purpose of this video is to show
how the EGG is used both for research and in the clinic. The principles of
operation will be described along with a general description of how the device
is positioned and adjusted. We will also examine the parameters of the EGG
waveform and how the EGG has been used in the laboratory and in the clinic to
examine certain aspects of vocal function. The principle of
electroglottography in which the amplitude of an electrical carrier signal
is measured has been around for over 50 years. In 1940, Philip Faber published
an article on a new method for monitoring the arterial pulse frequency in
the area of the carotid artery. In 1957, Faber described the first application
of this technique for examining voice production. This videotape will
demonstrate two commercially available devices. The first system was developed
by Martin Rothenberg of Syracuse University and is marketed by Glottal
Enterprises. The other device was developed by Dr. Adrian Forsen and is
marketed by Laryngograph Ltd. All EGGs are comprised of two basic
components. First, there are the electrodes. The electrodes apply to the
neck as small, physiologically safe, electrical carrier signal in the 1 to 5
megahertz range. These wires connect the electrodes to a second component, the
processor box. The processor contains electronics that generate the carrier
signal and that measure small modulations in that carrier signal that are caused
by the voice. These modulations in the carrier amplitude are what we define as
the EGG signal. The box also contains filters, amplifiers, and other components
that optimize the signal for display and for analysis. To attach the EGG to a
subject, I'll position the electrodes on the neck, one on each side of the
thyroid cartilage. To ensure good contact between the electrodes and the skin, a
small amount of water-soluble electrode paste is applied to each of the
electrodes. The electrodes are first positioned on the neck in the general
region of the vocal folds and can be held in place with the two fingers or
with this strap. I'll use this strap here.
Good. The position of the electrodes relative to the vocal folds determines
the strength of the signal. Could you say ah? Ah. Okay. I'm moving the electrodes
up and down a bit until I get a good, strong signal. Try it once more. Ah. Once
more. Ah. That's good. EGG signal strength is monitored on the meter of the device,
and I can also use an oscilloscope to determine the position that yields the
greatest output signal. A small percentage of subjects' EGG outputs are
too weak to be usable. The strength of the EGG signal is determined by the
person's neck anatomy. Those subjects having thinner necks and more prominent
thyroid cartilages tend to have stronger EGG signals. The signal strength meter
on this device can be useful in determining whether the EGG signal can be
relied on as a measure of vocal fold contact. The amount of vocal fold contact
occurring during each glottal cycle changes cyclically. This slow motion film
shows the folds alternating between being fully separated and fully together.
Notice that after the folds first make contact, they continue to approximate
along their vertical dimension and then begin to separate. These cyclic changes
in contact area between the vocal folds occur from the bottom to the top of the
device, resulting in the undulating movement that characterizes vocal fold
vibration. The cyclic variations in glottal area are shown in this typical
trace. We can see the relationship between changes in glottal area and the
other signals we've examined by adding them to the display. In the trace polarity
shown here, upward movement of the EGG trace indicates an increasing amount of
vocal fold contact, while a downward deflection of the trace shows diminished
vocal fold contact. This type of display is called the vocal fold contact area
signal, or VFCA. It is sometimes easier to interpret the signal if it is inverted
so that the top represents less vocal fold contact and the bottom represents
more contact. This orientation makes it easier to compare the signal to the flow
glottogram, for example. When the trace is inverted, we refer to the display as
the inverse vocal fold contact area signal, or IVFCA. A switch on this device
permits the user to choose either polarity. We can divide the EGG signal
into a few time phases corresponding to when the vocal folds are apart,
approximating, or separating. Notice that during the time the folds are approximated,
the signal is changing. This is because the amount of vocal fold contact
continues to increase after initial contact is made. Contact area increases
until the folds are most approximated, and then the amount of contact begins to
decrease. Also notice the flatter portion of the trace when the vocal folds are
not in contact. Although the distance between the folds is changing, there is
very little change in the amount of contact between them. Since it is
primarily the contact between the vocal folds that affects the electrical
resistance seen by the electrodes, the EGG signal changes very little during
this portion of the cycle. Thus, there is little information conveyed by the EGG
signal about the open portion of the vibratory cycle of the vocal folds other
than the duration of this period. The EGG has been primarily used in the
laboratory. Perhaps its most common use has been for detecting vocal fundamental
frequency. It's a good signal for this purpose because it has a strong
fundamental frequency component, thus permitting determination of fundamental
frequency using a simple threshold, crossing, zero, crossing, or peak-picking
schema. Unlike acoustically-based strategies for extracting fundamental
frequency, the EGG does not change significantly for different vowels, thus
providing a much more constant waveform for extracting fundamental frequency.
The EGG signal has also been used to examine the extent of abduction or
abduction during voicing. A person speaking with a breathy voice quality
often has EGG signals whose contact duration, the part of the glottal cycle
during which the vocal folds are together, is greatly reduced as compared
to normally voiced productions. The systematic changes in vocal fold contact
associated with the overall abduction or adduction of the folds has been
examined for stutterers, for deaf children, and for normal hearing speakers.
An additional feature of this EGG system permits fairly precise positioning of
the electrodes and monitoring of laryngeal height. Each electrode disc is
actually a pair of electrodes, one above the other. This results in two EGG
signals being generated. When the amplitudes of both signals are equal, it
means that the electrodes are positioned vertically with the vocal folds exactly
between the two pairs. This can be useful for positioning the EGG on different
subjects in approximately the same vertical location. An additional advantage
to this twin electrode arrangement is that as laryngeal height changes during
voicing, the relative amplitudes of the top and bottom pairs of electrodes will
change. The ratio of signal amplitudes is converted by the processor to a signal
that monitors laryngeal height. This specially designed calibration device
permits quantifying the laryngeal height signal in centimeters for each subject.
The laryngeal height feature has been used to examine the singing voice and
deaf speakers' control of voice. Let's now look at the other EGG system shown
earlier, the laryngograph. This device is available with a computer interface
card and a set of programs that enable display and analysis of the EGG signal
using a PC computer. One of the displays in the software package provides a
visual representation of fundamental frequency obtained from the EGG signal
against time. The split screen displays a template or model on the top and
provides the learner with feedback of the fundamental frequency of their attempt.
This type of display is useful for both deaf clients and for those learning a
second language. A second display permits analysis of the speaker's fundamental
frequency characteristics during connected speech. After acquiring the EGG
signal for a period of time, the software displays a histogram of the fundamental
frequency distribution together with statistics that characterize the
distribution. This feature is useful for diagnosis of fundamental frequency
related voice problems and for monitoring changes in average fundamental
frequency associated with therapy. The next display shows the actual EGG
signal either alone or in combination with other signals, such as the acoustic
signal from the microphone shown here on the top trace. This provides a means for
both the researcher and clinician to examine changes in global waveform
characteristics associated with different vocal pathologies or disorders.
For example, it is useful for examining and quantifying EGG attributes
associated with production of breathy voice productions or associated with
articulatory gestures of the vocal folds for consonant de-voicing. In this brief
program, we've seen that the electroglottograph provides a potentially useful way
of viewing vocal structures. The changes in vocal fold contact viewed either
alone or in combination with other signals, such as the flow glottogram, can
provide useful information about fundamental frequency, vocal fold contact,
and laryngeal height. While the device has limitations, it can prove a useful
supplement to existing research and clinical instrumentation.
The human voice is a unique instrument that serves basic needs of communication.
While we are used to listening to the voice and can easily make acoustic
measurements of the voice source, actually observing the vibrations is
difficult due to the inaccessible position of the larynx. Hence, special light
equipment and techniques are needed in order to visualize what's going on in the
larynx. One piece of equipment is the rigid endoscope. It's a cylinder with an
optical system inside. At one end is a lens that allows viewing objects at an
angle to the endoscope. At the other end is an attachment for a camera and a
light source. The endoscope is attached to the light source and to the camera.
The tongue is pulled forward during the recording to get a clear view of the
glottis. Recording laryngeal vibrations using a rigid endoscope thus prevents
normal articulation and can only be used during sustained phonation. For clinical
examinations, this is usually sufficient. The view through the endoscope shows the
interior portion of the larynx at the bottom of the screen. At the top of the
screen are the two arytenoid cartilages that move to adduct and abduct the vocal
folds. The white spot on each vocal fold close to the arytenoid cartilage is the
vocal process. In normal light, the view of the vibrating glottis is blurred.
No more pitch.
High pitch.
In scopic light, the individual vibrations can be observed.
Another instrument for laryngeal observation is the flexible fibroscope. It
consists of a bundle of thin glass fibres that transmit light and images. The
fibroscope can be inserted through the nose and thus does not impede normal
articulation. The fibroscope is connected to the light source and to the camera.
Before the insertion, an anesthetic and decongestant can be sprayed into the
nasal cavity.
During breathing, the glottis widens during inspiration and narrows during
expiration.
The mode of the vibrations is controlled by varying the degree of adduction of
the folds, less adducted for breathy voice and more adducted for pressed voice.
The fundamental frequency of the voice is controlled by changing the tension of
the folds. The length of the folds increases with tension and fundamental
frequency.
During speech, the voice source must be turned off for the production of
voiceless consonants. This is done by opening and closing the glottis at the
appropriate times. E P, E B, E F E, E C, E E.
An outside view of the larynx shows light passing into the trachea when the
glottis opens between phonations.
A light sensor placed on the neck below the glottis converts variations in light
passing through the glottis into an electric signal. This technique is called
transillumination.
The insert shows the output from the light sensor. Increasing glottal opening is
displayed upwards.
E P, E B.
The transillumination technique is particularly useful for recording glottal
adduction and adduction during voiceless consonants. Here it provides information
on lifting of the glottis and gesturing. For obvious reasons, the signal cannot
easily be calibrated.
E he, E P, E B, E F E, E V, E C,
E he, E P, E B, E F E, E V, E C,
E C.
Transillumination offers a useful tool for studying the coordination between
oral and arrhythmic articulatory movements. In this experiment, transillumination
is used to record glottal movements. Small, light emitting diodes are placed on
the lips and on the jaw splint to record movements of oral articulators. The
diodes are tracked by a special camera. This experiment also involves applying a
mechanical load to the lower lip during speech. A paddle rests on the lower lip.
The paddle is connected to a motor that can be activated to pull the lip
downwards. The motor is operating under computer control so that the load can be
applied to the lip when the subject is making the oral closure for a belabial
stop.
The rationale for this research is that the nature and time course of the
responses to the load can reveal the motor organization and reflex structure of
the act.
The view of the glottis is monitored during the experiment to make sure that
there is a free passage for the light from the fibroscope to the glottis.
Studies using this experimental paradigm show that when the lower lip is
perturbed, the upper lip and the jaw compensate for the perturbation so that
the bilabial closure is still achieved. However, it appears that the timing
between the larynx and the upper articulators is affected by such a
perturbation.
Go. The south back boat.
Go. The south back boat.
No perturbation on that jaw.
Go. The south back boat.
Go. The south back boat.
Go. The south back boat.
Go. The south back boat.
Go. The south back boat.