Neural correlates of binaural spectral and temporal fusion

About This Grant

Project Summary This project seeks to understand the neurophysiological basis of binaural fusion in the auditory system. When interacting with the world, humans and other animals must process sound information arriving from multiple different sources. Depending on the location of a sound source, the signals reaching the two ears will differ, and the auditory system uses these binaural cues to determine sound location. When multiple sound sources are present, the auditory system also uses binaural information to determine if sound features belong to the same or different source. When sounds reaching the two ears have similar spectral and temporal properties, they are typically fused and perceived as belonging to a single source. In patients with hearing loss, binaural fusion can occur over a wider range of binaural differences, and multiple sources can be incorrectly fused and perceived as belonging to a single source. Using a combination of behavioral and neurophysiological approaches, this study will determine the conditions under which humans and animals experience binaural fusion and the neural mechanisms that support these percepts. A critical aspect of this work that enables linking human and animal behavior is the development of new, objective behavioral measures of binaural fusion. The first set of experiments will use this objective behavioral approach to determine factors that lead to fusion of binaural spectral information in humans and animals. It will then use high-channel count, single-unit neurophysiological recordings to characterize the neural representation of fused and segregated stimuli. The second set of experiments will use a similar approach to determine neural correlates of fusion of temporal sound modulations. The third, and final experiments will measure the neural response to large sets of natural binaural stimuli. It will then fit a deep learning-based encoding model to describe the neural response to any arbitrary binaural sound. This model will be validated against its ability to predict neural responses to fused and non-fused stimuli. It will then be used to synthesize novel stimuli that are predicted to produce a fused percept in human subjects. Understanding the conditions under which fusion occurs and the neurophysiological mechanisms that mediate fusion will provide insight into perceptual deficits following hearing loss and can be used to guide the development of new treatments that reduce abnormal fusion.