CORTICAL REPRESENTATION OF AUDITORY SPACE

点击翻译按钮获取中文摘要

The objectives of the proposed research are to identify specific elements of sound spectra that the auditory system uses to identify the location of a sound source and to explore the cortical code for sound location. The directional transfer functions of the external ears shape the spectra of sounds, thereby providing cues to the vertical and front/back locations of sounds. Sharp or notches introduced into a stimulus spectrum result in spatial illusions, which are evident as systematic mis-localizations. Behavioral experiments in humans, combined with computational modelling, will dissociate specific elementary fares of spectral peaks and notches (e.g., energy maxima or minima, spectral slopes, rapid changes in slopes). This approach will place constraints on neuronal models of localization mechanisms. Determined efforts by several groups have failed to find a topographic map of auditory space in the cortex, at least not in the conventional sense of an array of neurons that vary in spatial tuning as a function of cortical place. Recent results from this laboratory, however, have demonstrated that the temporal firing patterns of single auditory neurons in the cat's anterior ectosylvian area (area AES) can code the azimuth of a sound source. Proposed neurophysiological experiments in cats will compare location ceding in areas AES, A1, and A2. Experiments will focus on whether particular neurons code localization per se, or whether they merely are sensitive to particular localization cues. Toward that goal, the coding of sound source elevation will receive particular attention. Additional neurophysiological experiments will explore parallels between the behavior of human listeners and the physiology of cat cortical neurons. Human localization behavior tolerates considerable variability in the shapes of source spectra, but fails when spectra contain sharp peaks or notches. Temporal coding of stimulus locations by cortical neurons will be tested with stimuli that vary in spectral shape. To the degree that particular cortical neurons form part of the substrate for sound localization, one might expect cortical location coding to succeed and fail under conditions analogous to those observed in human behavior. The auditory cortex is a key element of the temporal lobe, which is the cortical substrate for many communicative processes, is a common site of epileptic foci, and can be involved in schizophrenia. The proposed neurophysiological experiments address a fundamental issue of sensory coding in the cortex, the issue of place coding versus temporal coding. The spectral shape experiments explore basic sound localization mechanisms and, by testing for parallels between cortical ceding and behavior, provide a test of the behavioral significance of the cortical temporal coding results. An understanding of basic cortical coding mechanisms should contribute to the valuation of temporal lobe pathology and the design of therapeutic responses to injury and disease.