A major issue in sensory systems processing is how a relevant signal is extracted from distracting background noise in the environment. Hearing is excellent for detecting certain highly relevant signals, such as communication signals, and we know little about the neural mechanisms involved. This project uses a combination of neurophysiological recording with computational modeling to determine how single cells can detect features of an auditory signal. Computational techniques are used to change particular features of the sounds presented to the ear, and to analyze the responses of the cells. Spatial and temporal patterns of activity from single cells are compared to the patterns of activity coming into those cells from auditory nerve fibers, in a part of the brainstem called the anteroventral cochlear nucleus. These experiments test hypotheses about how much "feature detection" can be done at this level of the auditory system, and how much of this depends on analysis of the "spectral shape" of combined frequencies in a complex signal. Results from this study will be important beyond auditory perception to studies of sensory processing in general, to understanding how the brain integrates a wide range of simultaneous activity, and to better design of electronic devices for artificial listening and auditory aids.
