2007; A Physiologically-Based Population Rate Code for Interaural Time Differences (ITDs) Predicts Bandwidth-Dependent Lateralization

Abstract 1062, Date 1:00 pm, Monday, February 12, 2007 (24 hours) Session K12: Poster
A Physiologically-Based Population Rate Code for Interaural Time Differences (ITDs) Predicts Bandwidth-Dependent Lateralization
*Kenneth E. Hancock
ITDs are encoded centrally in the medial (MSO) and lateral (LSO) superior olives. It is a longstanding view that these neurons are conceptually arranged in an array with characteristic frequency (CF) on one axis and best ITD (BD) finely distributed on the other to form a labeled-line code of ITD. One important perceptual aspect of binaural hearing that has been explained using labeled-line models is the dependence of lateralization on stimulus bandwidth. The labeled-line model has been challenged by physiological data from the guinea pig, confirmed in the cat and gerbil, showing there is not a full complement of BDs for all CFs. It has been suggested that ITD may be represented by a rate code, in which the activity of many neurons pool to form a single ITD channel on each side of the brain. We demonstrate the ability of such a model to predict bandwidth-dependent lateralization. Each channel comprises 625 model neurons, whose CFs and BDs are distributed to mimic physiological data from the cat inferior colliculus. Each neuron cross-correlates the bandpass-filtered stimuli to the two ears, and the channel output is simply the summed firing rate of its neurons. The model predicts lateralization as a function of bandwidth, if four channels are used, one representing each MSO and one each LSO. The lack of need for labeled lines minimizes the precision required in the BD array and may explain how a viable ITD representation is formed despite the existence of several factors that determine the BD of individual neurons. Finally, the outputs of the population rate model are appropriate for controlling muscle movements that orient the head toward a sound source. Thus, it is conceivable that the neural ITD code is an evolutionary remnant of a primitive coupling between sensory stimulation and motor response.