Physiological modeling for hearing aid design

Physiological data from hearing-impaired cats suggest that conventional hearing aid signal-processing schemes do not restore normal auditory-nerve responses to a vowel [Miller et al., J. Acoust. Soc. Am. 101, 3602 (1997)] and can even produce anomalous and potentially confounding patterns of activity [Schilling et al., Hear. Res. 117, 57 (1998)]. These deficits in the neural representation may account at least partially for poor speech perception in some hearing aid users. An amplification scheme has been developed that produces neural responses to a vowel more like those seen in normal cats and that reduces confounding responses [Miller et al., J. Acoust. Soc. Am. 106, 2693 (1999)]. A physiologically accurate model of the normal and impaired auditory periphery would provide simpler and quicker testing of such potential hearing aid designs. Details of such a model, based on that of Zhang et al. [J. Acoust. Soc. Am. 109, 648 (2001)], will be presented. Model predictions suggest that impairment of both outer- and inner-hair cells contribute to the degraded representation of vowels in hearing-impaired cats. The model is currently being used to develop and test a generalization of the Miller et al. speech-processing algorithm described above to running speech. [Work supported by NIDCD Grants DC00109 and DC00023.] a)Now with the Dept. of Electrical and Computer Engineering, McMaster Univ., 1280 Main St. W., Hamilton, ON L8S 4K1, Canada.