The long range goal of this work is to understand the encoding and processing of human speech by the brain.
The specific aims of this proposal focus on understanding the role of efferent systems, specifically the olivocochlear system, on stimulus encoding and processing in the awake, behaving cat. Although these efferent systems have been shown to be active in the anesthetized preparations used previously, recent studies in this laboratory on behaving cats have shown enhancements in the neural representation of sounds in awake cats relative to those in anesthetized cats. Rate level functions for best frequency tones for auditory-nerve and cochlear nucleus units are compressed by background noise and their dynamic ranges shifted with respect to those measured in quiet. In awake cats compression is reduced and dynamic range shifts increased for cochlear nucleus units in comparison with those previously measured in anesthetized cats. These changes may improve the ability of the cat to discriminate sounds in noisy environments. The evidence suggests that these differences may be mediated by the olivocochlear bundle (OCB) and the experiments proposed here all explore that suggestion. Rate-level functions for best frequency tones in quite and in noise will be measured in single cochlear nucleus units in awake cats performing a simple discrimination task; within-cat comparisons will be made between three different physiological states: awake with OCB intact and lesioned, and anesthetized. Representations across populations of cochlear nucleus units of vowel spectra and tone frequency in quiet and noise will be measured in cats that are discriminating between two stimuli (two vowels of slightly different formant structure, or two tones of different frequency). The representations in cats with intact and lesioned OCB will be compared with those in anesthetized cats. Quantitative comparisons will be made on the basis of models that use the neural representations to compute predictions of discrimination performance. The results have direct relevance to the design of prostheses for the hearing impaired, including cochlear and CNS prostheses and hearing aids. The are also important to a broad range of aids for the hearing impaired that require computer recognition of speech, such as speech-training aids.
