Wednesday, March 20, 2002 3:05 pm LCB 323
Title: Circuits for spatial integration of information in the primate visual cerebral cortex.
Abstract - At the early stages of visual processing an image is decomposed into an ordered collection of elementary features. Thus, for example neurons in the primary visual cortex respond to the specific orientation of line segments, their direction of motion or color, within a circumscribed region of visual space, called the cell's receptive field (RF). A central question in visual neuroscience is how such local information is integrated to generate a unitary and coherent percept. Feedforward models see visual information ascending through a hierarchy of visual cortical areas, with cells at each successive stage processing information from larger region of space, and responding to more complex aspects of a visual stimulus. Despite some supporting evidence, these models alone cannot account for object recognition, especially in very "crowded" visual scenes. Additionally, there is a great deal of experimental evidence suggesting that spatial integration of information across distant parts of the visual scene can occur at very early stages of processing. Feedforward models and classical receptive field concepts cannot explain such phenomena. The work I will present is aimed at identifying the cortical substrates underlying long-distance signal integration in the early visual cortex of primates. Our main conclusions provide an alternative to the feedforward model, suggesting an important role for feedback circuits from higher to lower cortical areas in global spatial integration of visual informaton.
For more information contact J. Keener, 1-6089