Picture ErrorPicture ErrorPicture ErrorPicture ErrorPicture Error

Kresge Hearing Research Institute

Department of Otolaryngology

Central Systems Laboratory

Central Systems Menu

News

Central Systems Laboratory

Shigeto Furukawa

Postdoctoral Research Fellow
Email: shig@umich.edu
Phone: (734) 764-5167
Fax: (734) 764-0014

Accomplishments

Hobbies

Research Project

Localization of sound source is an important function of the auditory system, and intact auditory cortex is essential for normal sound localization. Our previous studies have shown that the spike patterns of single neurons in the auditory cortex can carry information about sound-source locations throughout 360� of space, and such �panoramic� neurons are distributed widely throughout the auditory cortex. The results have lead us to an alternative �distributed code� in which information about any point in the auditory space is distributed across large populations of neurons. The present project deals with questions as to how populations of cortical neurons, or neural ensembles, work for location coding. Our hypothesis is that an ensemble of neurons is more than a simple sum of individual neurons: Activities of cortical neurons are coordinated to code sound-source location. Specific aims include 1) estimating the amount of information about sound locations carried by ensembles of cortical neurons, and 2) exploring specific information-bearing features of the ensemble spike patterns, particularly in terms of relative activity patterns between neurons.

With the help of thin-film silicon recording probes ("Michigan Probe"), we are able to record single units from the cat's auditory cortex at as many as 16 sites simultaneously. An artificial neural network (ANN) algorithm is used to identify sound-source locations by recognizing characteristic spike patterns of neural ensembles elicited by sounds at various locations. We can interpret the accuracy of ANN estimations as an empirical measure of the amount of stimulus-related information carried by the spike patterns. So far, we have applied those techniques for analyzing neural responses of neuron pairs and of neural ensembles.

Related Publications