Spatial hearing provides a powerful mechanism for discriminating sounds in noisy backgrounds. Project 1 will explore cortical mechanisms of spatial hearing, first in quiet, then in ambient noise. An audible sound must result in a time-varying distribution of activity across the cortex, which will be referred to here as the """"""""cortical image"""""""" of that sound. To the extent that specific cortical images are unique to particular sounds, sound sources can be identified by their cortical images. This project will use multi-channel recording techniques to sample single-unit activity within cortical images at up to 32 sites simultaneously. In the context of this application, a cortical image may be regarded as the neuron-level correlate of the functional images that are observed with MRI, PET, and functional optical imaging, in this case with single-cell spatial and 1-ms temporal resolution. The goal of Specific Aim 1 is to identify the features of cortical images that can carry stimulus-related information. Artificial neural networks will be used to identify sound-source locations on the basis of cortical responses, and the accuracy of identification will reveal the relative amount of information that can be carried by particular elements of the responses of single neurons and of ensembles of up to 32 neurons. Particular study will be given to the relative amplitude and timing of activity among multiple units.
Specific Aim 2 is designed in parallel with Project 2 to study spatial hearing in ambient noise. The degree to which specific elements of cortical responses track behavioral sensitivity to masker levels will reveal the relative significance of those features for signal detection and discrimination in ambient noise. The auditory cortex is a key element of the temporal lobe. The proposed project addresses the issue of how stimuli might be coded in the cerebral cortex by the coordinated activity of populations of neurons. The studies of the cortical image of a sound will complement diagnostic functional imaging techniques at the level of single neurons. This work should contribute to our understanding of the neural coding of perception, the evaluation of temporal lobe pathology and to the design of therapeutic responses to injury and disease.
Showing the most recent 10 out of 54 publications
