-
Andrew J. Viterbi Distinguished Lecture with Dr. Toby Berger
Thu, Mar 09, 2006 @ 05:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
"Information Exchange Among Neurons in Sensory Cortex"PROF. TOBY BERGERProfessor of ECE, University of VirginiaProfessor Emeritus, Cornell UniversityReception at 5:00p.m.Lecture at 6:00p.m.Abstract:Each of the tens of millions of neurons in primate sensory cortex receives information in the form of neural spike trains from some 10,000 other such neurons, referred to collectively as its "afferent cohort". The axon of each neuron in the afferent cohort of neuron N is incident to a synapse on N's dendritic tree. In response to this "bombardment," neuron N generates a spike train of its own which propagates along its axon to the some 10,000 neurons in its afferent cohort. The nature of this "neural code" and of the information that it does and does not communicate has long been a subject of intense study by both theoreticians and experimentalists. For several decades much of this debate was concerned with whether the neural code is a timing code or a rate code. Recent experiments have established that the firing patterns of individual neurons in primary visual cortex (i.e., brain regions LGN, V1, V2,...) are highly reproducible when stimulus intensities are made to vary nearly identically during each of a series of experimental trials. This finding speaks heavily in favor of timing codes being the dominant information-conveying mechanism in primary cortex.We propose and analyze a simple, biologically feasible model of neural coding in primary cortex in which the time variation in bombardment intensity that each neuron experiences gets accurately encoded into differential delays between the successive spikes it emits. That is, we proffer that neurons in primary sensory cortex exchange information via what communication theorists refer to as differential pulse position modulation (DPPM) and neuroscientists refer to as interspike interval (ISI) durations. This form of neural information exchange has several highly desirable properties including low delay, low complexity, high throughput, and energy efficiency.Three sources of error in the conveyance of data via ISI durations are analyzed: (i) timing jitter in spike generation instants (ii) timing jitter engendered by differential rates of axonal propagation of successive spikes, and (iii) timing jitter in the accuracy with which spike arrival times can be estimated and remembered at synapses. The analysis reveals that neural spiking thresholds must decay with time; preliminary experimental results suggest that this is indeed the case.The mathematical tools employed include Stein-Chen Poisson approximation theory, analysis of the variance of the threshold crossings times of filtered Poisson processes, and classical analysis of the estimation of the arrival time of a known pulse shape corrupted by additive Gaussian noise.Bio: Toby Berger was born in New York, NY on September 4, 1940. He received the B.E. degree in electrical engineering from Yale University, New Haven, CT in 1962, and the M.S. and Ph.D. degrees in applied mathematics from Harvard University, Cambridge, MA in 1964 and 1966, respectively. From 1962 to 1968 he was a Senior Scientist at Raytheon Company, Wayland, MA, specializing in communication theory, information theory, and coherent signal processing. From 1968 through 2005 he was a faculty member at Cornell University, Ithaca, NY where he held the position of Irwin and Joan Jacobs Professor of Engineering. In 2006 he became a professor in the ECE Department of the University of Virginia, Charlottesville, VA. Professor Berger's research interests include information theory, random fields, communication networks, wireless communications, video compression, voice and signature compression and verification, neuroinformation theory, quantum information theory, and coherent signal processing. He is the author of Rate Distortion Theory: A Mathematical Basis for Data Compression and a co-author of Digital Compression for Multimedia: Principles and Standards, and Information Measures for Discrete Random Fields.HOSTED BY THE DEPARTMENT OF ELECTRICAL ENGINEERINGCelebrating 100 years of Engineering!For more information: http://csi.usc.edu/news/viterbi-invitation.pdf
Location: Ethel Percy Andrus Gerontology Center (GER) - ontology Auditorium
Audiences: Everyone Is Invited
Contact: Rosine Sarafian