The mammalian cerebral cortex is characterized by multiple cortical areas, each with a separate representation of the sensory world. However, the role of multiple representation in sensory information processing is unknown. As a prelude to understanding how multiple cortical areas cooperatively process visual percepts, this proposal aims to study contour representation within V2 (the second visual cortical area) and its interaction with contour representation in V1 (primary visual cortex). Orientation selective cells in V1 are well known to be responsive to luminance contrast edges. However, many visual contours are not defined by luminance contrast but rather by higher order features (such as occluded contours or texture pattern borders). Area V2, the second major stage of visual cortical processing, is characterized by cells which prefer such higher order features. This proposal will explore the idea that different cortical areas process different levels or abstractions of visual contours and examine how these separate representations interact. Specifically, the organization and interactions between cells responsive to real contours and those responsive to higher order contours will be examined. These circuitries will be examined for each of the color, form, and disparity domains which together comprise V2. The stated goals will be achieved by using a combination of optical imaging, electrophysiological, and anatomical methods. V1-V2 interactions will be studied with both cross correlation methods and optical imaging of V1-V2 activation in response to """"""""spot"""""""" stimuli (a functional tract tracing method); these will be compared with connectivity patterns revealed anatomically. In addition, differences in contour saliency due to geometrical or featural characteristics will he related to stimulus- related changes in interaction patterns. By elucidating the neural circuitry underlying contour perception in V1 and V2, this research will offer insight into whether different cortical areas mutually reinforce the same information (i.e. interactions between similarly oriented real and higher order contour cells) or whether a more hierarchical relationship exists in which the outputs of one area are used to construct a quite different representation in another area (interactions between differently oriented real and higher order contour cells). Alternatively, different strategies could be used under different contexts, a possibility that would be supported by stimulus-related changes in connectivity or activation of different populations of contour cells in V2.
Showing the most recent 10 out of 29 publications
