Michael Epstein

On using tone-burst otoacoustic emissions to measure basilar-membrane response in humans

Date: Thursday January 22, 2004

Abstract: Otoacoustic emissions are sounds produced by healthy cochleae in response to stimulation. They are byproducts of the activity of outer-hair-cell amplification. Because cochlear response maps frequencies to places, using a narrow-band signal as a stimulus makes it possible to examine the activity of a particular location on the cochlea. Therefore, tone-burst otoacoustic emissions (TBOAEs) are a potentially useful tool for examining basilar-membrane activity in humans. Knowledge of basilar-membrane activity is essential to the understanding of the perception of sounds, especially loudness. However, basilar-membrane activity cannot be observed directly in humans. Because TBOAEs arise on the basilar membrane, their amplitude is likely to be closely related to the motion of the basilar membrane itself. The use of this metric requires knowledge of just how the relation between TBOAEs and basilar-membrane input-output (BM I/O) activity depends on measurement and analysis parameters. Before implementing a TBOAE procedure, it is necessary to examine experimental paradigms for measuring TBOAEs. Three parameters will be examined:
1) response-window timing
2) response delinearization
3) small stimulus-frequency changes.

A judicial choice of these experimental parameters will optimize the assessment of BM I/O functions with TBOAEs. If these parameters do have a profound effect on TBOAE measurements, it may not be possible to determine which parameters, if any, provide the most appropriate view of the BM I/O function.

Once an appropriate method for measuring TBOAEs has been selected, comparisons will be made with known psychoacoustic measures of BM I/O. These include:
1) pulsation threshold
2) loudness from temporal integration,
3) loudness matches using multitone complexes
4) the growth of forward masking.
In addition, a rough comparison between measured TBOAEs and non-human physiological data (from other experimenters) will be made.

TBOAEs are likely to reveal BM I/O more quickly and reliably than psychoacoustical measures and therefore, may be an excellent method for examining basilar-membrane activity noninvasively. This knowledge may help provide a very valuable, objective measure of the relationship between physical sound intensity and loudness.

Thesis Committee:
Prof. Søren Buus (advisor) - ECE
Prof. Dana Brooks - ECE
Prof. Mary Florentine - Speech-Language Pathology and Audiology