Development and real-time implementation of a wavelet based speech processing strategy with noise suppression for cochlear implants

Cochlear implants are prosthetic devices used to provide partial hearing in profoundly deaf people. The performance of cochlear implants deteriorates in noisy environments compared to quiet conditions. This reduction in performance has been attributed to various factors which include lower spectral and temporal resolution and inadequate noise suppression. In the first part of this dissertation, a wavelet-based speech processing strategy for cochlear implants is introduced. Three wavelet packet decomposition tree structures are considered and their performance in terms of computational complexity, spectral leakage, and fixed-point accuracy are compared to other commonly used strategies in cochlear implants. A recursive method of updating the wavelet coefficients is introduced to provide real-time processing. It is shown that the proposed strategy achieves higher analysis rates and a lower amount of spectral leakage than the existing strategies. In the second part of this dissertation, an adaptive cochlear implant system which is capable of classifying the background noise environment in real-time for the purpose of adjusting or tuning the noise suppression algorithm to that environment is introduced. The tuning is done automatically with no user intervention. Objective quality measures are used to show the superiority of this adaptive system compared to a conventional fixed noise-suppression system. Steps taken to achieve the real-time implementation of the entire system, incorporating both the cochlear implant speech processing and the background noise suppression, on the FDA approved PDA research platform for cochlear implant studies are reported along with the timing results.