Thesis (Ph. D.)--Harvard-MIT Division of Health Sciences and Technology, 2004. / Includes bibliographical references (p. 191-203). / The importance of the cochlear active mechanism in the reception of speech in different types of noise was explored. The perceptual effects of loudness recruitment, a consequence of loss of the active mechanism, were assessed in simulated-loss listeners using a multiband expansion algorithm that models abnormal cochlear linearity. While this algorithm, which derives the expansion characteristic from absolute hearing thresholds, can accurately simulate the mean speech intelligibility results of hearing-impaired listeners, its ability to simulate the performance of individual listeners is limited. Given the relationship between loudness perception and the active mechanism, deriving the expansion characteristic from estimates of cochlear compression should provide a more accurate model of an individual listener's impairment. Towards this aim, the reliability of two psychoacoustic methods used to estimate the magnitude of compression (growth of masking and temporal masking) was assessed. Results suggest that growth of masking is a more reliable measure of compression in listeners with both normal and impaired hearing. The relationship between the compressive characteristics of the auditory system and speech perception in complex acoustic backgrounds was also evaluated. The operational status of the active mechanism was assessed behaviorally using three independently derived measures: (1) slope ratio of off- and on-frequency growth of masking functions, (2) equivalent rectangular bandwidth of auditory filters, and (3) masker-phase masking differences. These measures were correlated with speech reception thresholds (SRTs) in backgrounds with and without spectral and/or temporal-modulations. The relationship between slope ratios, filter bandwidths, and the maximum / (cont.) SRT difference (SRT in steady noise minus SRT in temporally modulated, spectral gap noise) was significant. These results indicate that the ability to take advantage of momentary fluctuations in the amplitude or frequency spectrum of background noise requires an intact active mechanism. The speech reception performance of two hearing-impaired listeners was modeled using a customized version of the expansion algorithm. The algorithm was customized to an individual's impairment based on psychoacoustic measures used to evaluate the integrity of the active mechanism. The maximum SRT difference in the simulated-loss listeners more closely matched the results of their hearing-impaired counterparts, compared to SRTs measured using the original algorithm. These results provide further evidence of the importance of the active mechanism to the perception of speech in modulated noise. / by Peninah S. Rosengard. / Ph.D.
| Identifer | oai:union.ndltd.org:MIT/oai:dspace.mit.edu:1721.1/28598 |
| Date | January 2004 |
| Creators | Rosengard, Peninah S., 1970- |
| Contributors | Louis D. Braida and Andrew J. Oxenham., Harvard University--MIT Division of Health Sciences and Technology., Harvard University--MIT Division of Health Sciences and Technology. |
| Publisher | Massachusetts Institute of Technology |
| Source Sets | M.I.T. Theses and Dissertation |
| Language | en_US |
| Detected Language | English |
| Type | Thesis |
| Format | 203 p., 9461477 bytes, 9489930 bytes, application/pdf, application/pdf, application/pdf |
| Rights | M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission., http://dspace.mit.edu/handle/1721.1/7582 |
Page generated in 0.0022 seconds