| |
IVANS
Measures
IVANS provides 47 acoustic voice measures including all the familiar clinical
measures plus advanced clinical measures that are highly sensitive and relevant
but have never been available to clinicians. IVANS acoustic parameters include:
- Instantaneous and mean Jitter and Shimmer
- Amplitude Perturbation Quotient (APQ)
- Pitch Perturbation Quotient (PPQ)
- Harmonic to Noise Ratio (HNR)
- Normalized Noise Energy (NNE)
- Vocal Tremor
- Spectral Flatness Ratio (SFR)
Informative data displays are available for all measures, and for a composite
profile of key measures. Additional interactive displays are provided for the
pitch track and formant tracks. Additional forms facilitate the systematic recording
of perceptual/voice quality ratings.
Acoustic Measures of Voice Health
IVANS provides comprehensive support for vocal function analysis based on objective,
noninvasive acoustical measures. Voice samples are recorded directly to computer
disk, where they may be stored as distinct files. These samples are then available
for a wide range of acoustic analyses, and also for perceptual ratings. They can
also be archived and recalled at a later date to evaluate client progress.
IVANS supports CD-quality (44.1 kHz, 16 bits/sample) recording and playback on
all Windows-compliant 16-bit audio systems. The recording and playback of digital
signals is done directly to and from disk, so signal length is limited only by
hard disk space. Entire signals or selected parts of any signal can be played
back from all relevant analysis forms, giving you added confidence in the samples
you are measuring.
IVANS provides both familiar measures of voice performance, and many advanced
and previously-unavailable instrumental measures. Because voice fundamental frequency
is the key to accurate measurements for many voice parameters, IVANS provides
three advanced methods -- all extensively tested for use with disordered voices,
as well as with normal voices -- for measuring voice pitch [1].
In implementing IVANS, much attention has been given to the selection and careful
implementation of the acoustic measures. For example, in addition to providing
familiar measures such as "jitter" and "shimmer", calculated in several ways,
IVANS also implements advanced measures based on linear modeling of speech. This
approach has been shown to be particularly useful in the classification of vocal
pathology (e.g., [2]) but measures that use the approach have not previously been
available for clinical use.
Available voice measures include:
- Sustained Vowel Phonation (up to 15 vowel samples)
- Sustained S/Z Phonation (S/Z ratio is calculated automatically on multiple samples)
- Amplitude of Speech (mean RMS voice amplitude, in dB)
- Dynamic Range of Speech (amplitude difference, in dB, between the softest and
loudest utterances that a client can produce)
- Maximum Phonational Range (frequency difference, in Hz, between the lowest- and
highest- pitched utterances that a client can produce)
- Voice Fundamental Frequency (F0; rate of vocal fold vibration)
- Jitter (cycle-to-cycle variation in the rate of vocal fold vibration -- several
measures)
- Shimmer (cycle-to-cycle changes in the amplitude of the signal generated by the
vocal folds -- several measures)
- Glottal Noise (amount of unintended noise generated during voiced speech, both
HNR and NNE are obtained)
- Spectral Flatness Ratio (SFR)
- Vocal Rise Time (time required for a speaker to produce a tone of full amplitude)
- Vocal Fall Time (time required for a speaker to stop producing a tone)
- Voice Tremor (degree of variation in F0 or voice amplitude during sustained phonation)
- Spectral Analysis (reflecting vocal tract resonances), particularly Formant Center
Frequencies and Bandwidths
- Spectral Tilt
- Long Term Average Speech Spectrum
Samples may be recorded at the time you take the measures, or you may record the
samples and take the measures at a later time, after your client has departed.
Data are then displayed on the Vocal Function Test and Results Form, and can be
archived and later imported into reports, as desired.
Judgements of Voice Quality
Voice Quality Ratings
Voice Quality Ratings are systematic clinical judgements of specific characteristics
of a talker's speech. Clinical impressions of a client's speech can be recorded
in terms of these characteristics, using a seven-point rating scale. At any point
in the rating process, a speech sample can be selected and played to assist in
the rating process. Rating data are then stored in the client's file and can be
recalled later for inclusion in a report, or for comparison with similar evaluations
taken during another assessment of the client.
Characteristics that may be rated using IVANS include:
- Hoarseness
- Roughness
- Breathiness
- Tension
- Tremor Strain/ Stuggle Behavior
- Interruption of Voicing
- Diplophonia
- Speaking on Inhalation
Recording Subjective Perceptual Impressions
Another IVANS form lets you record your subjective perceptual impressions of the
client's voice. This allows specific voice characteristics that may be useful
in identifying specific voice problems to be recorded and saved along with the
client's instrumental data and other records. Such impressions may include:
- Aspects of the Pitch of the client's normal speaking voice, characterized in terms
of the appropriateness of the mean or average speaking fundamental frequency,
the presence or absence of frequency breaks, the appropriateness of the client's
phonational range, and the appropriate of perturbations of voice pitch
- Aspects of the client's Inflection, judged in terms of the appropriateness of
Inflection control and frequency variability
- Aspects of the rate & timing of vocal rise and fall time, of phonation time and
the presence of any voice stoppages
- Aspects of voice amplitude, including the appropriateness of the client's mean
overall speaking level, variability in amplitude, dynamic speaking range, and
the presence of any voice tremor, or other speech perturbation
- Clinical impressions of a client's sustained vowel performance.
Physical & Physiological Measures
To assist with record keeping, the tracking of client progress and the preparation
of reports, IVANS allows you to enter selected data that you have obtained using
non-acoustic, physical/ physiological instruments. Using this form, you can enter
Airflow data and EGG data (e.g., obtained with a Laryngograph or ElectroGlottograph),
as well as impressions observed from fiberoptic and/or stroboscopic examinations.
These data can then be stored in the file along with the IVANS acoustic measures
and other data for the client, and recalled when needed, for example to monitor
client progress or when preparing a clinical report.
References
Parsa, V. & Jamieson, D. G. (1999). A comparison of
high precision F0 extraction algorithms for sustained vowels. Journal of Speech
and Hearing Research, 42, 112-126.
Parsa, V. & Jamieson, D. G. Identification of pathological voices based on glottal
noise measures. Journal of Speech, Language, and Hearing Research, in press, to
appear in February 2000.
Order IVANS | Download
Demo |
|
