The binaural cues used to localize sound in the horizontal plane (i.e., azimuth) and the neural mechanisms by which these cues are encoded in the responses of neurons in the central auditory system are well known. What remains unclear is how the combined responses across neurons that are sensitive to ITD and ILD form a neural representation of azimuthal location that is robust to changes in natural auditory scenes. This proposal focuses on how the population response of neurons in the inferior colliculus (IC) encodes sound source azimuth with respect to two of these changes: 1) with changes in sound level, and 2) in the presence of a competing sound source.
The third aim i nvestigates changes in the population encoding of azimuth brought about by sensorineural hearing loss (SNHL).
These aims will be addressed by collecting azimuth tuning functions from single-unit recordings in the IC of awake rabbits. Population IC responses will be used to train and test a maximum-likelihood decoder of source azimuth as a means to both evaluate the azimuth encoding ability of population activity, and to test whether patterns of population activity specific to source azimuth are robust to changes in other sound parameters (e.g., sound level). Experiments in Aim 1 will use azimuth tuning functions collected at multiple sound levels to test whether performance of the decoder is invariant to sound level;or in other words, whether the location-specific patterns of activity across IC neurons (or a subpopulation of neurons) remain largely similar between sound levels. Experiments in Aim 2 will investigate the neural coding of sound location in a multisource environment with different sound source onsets. When the onset of one sound source precedes another, IC neurons may adapt to the early-onset source and """"""""unmask"""""""" azimuth information for the later arriving source. To test this hypothesis, azimuth tuning functions will be collected under target-alone and target-plus-interferer conditions (both simultaneous and onset difference). The decoder will be used to assess invariance of IC activity patterns when trained on target-alone data and tested on target-plus- interferer data. Finally, experiments in Aim 3 will test whether or not the azimuth encoding of IC neurons is degraded in rabbits with bilateral SNHL-the leading form of permanent hearing loss. SNHL will be induced in rabbits via high-level noise exposure. Following verification of permanent hearing loss two weeks after exposure, azimuth tuning functions will be measured in the IC. Decoder performance will be compared between noise-exposed rabbits and unexposed rabbits to evaluate whether there is a decrease in the ability of IC neurons to encode source azimuth following SNHL. Azimuth tuning will also be measured under conditions where only interaural time difference or interaural level difference varies with azimuth to assess any specific deficit i the neural representation of either binaural cue.
The experiments in the current proposal seek to understand how the brain represents the location of a sound source, particularly in everyday environments which contain multiple, overlapping sources of sound. The experiments also seek to isolate deficits in the neural processing of sound location of individuals with sensorineural hearing loss-a common form of hearing loss which is known to degrade sound localization ability.
| Day, Mitchell L; Delgutte, Bertrand (2016) Neural population encoding and decoding of sound source location across sound level in the rabbit inferior colliculus. J Neurophysiol 115:193-207 |
