Ventral and dorsal visual streams in action planning

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Scott H. Johnson, Ph.D. was trained as a cognitive psychologist and has established an active research program in object perception and prehension. The purpose of this award is to allow Scott to develop additional knowledge, skills and experience necessary to pursue both the behavioral and neural bases of these processes. This goal will be undertaken with the support and supervision of sponsor Scott G. Grafton and co-sponsors Nancy G. Kanwisher, Andrew Barto. Functional MRI, and psychophysical studies will take place at the the Center for Cognitive Neuroscience at Dartmouth College where Scott is currently a Research Assistant Professor. Additional training will be pursued through Dr. Barto?s computational motor lab at the University of Massachusetts and Dr. Kanwisher?s lab at M.I.T. The proposed research plan describes a series of studies designed to increase our understanding the contributions of the human ventral and dorsal visual streams to the planning of visually-guided actions; i.e., tool prehension and manipulation. These experiments employ a variety of strategies (event-related fMRI, psychophysics, and computational modeling) to evaluate hypotheses motivated both by computational considerations and what is currently known about the functional architecture of the visual and motor systems. It is well-established that extrastriate visual areas of primates are organized into two primary streams of processing: an occipital-temporal (i.e., ventral) stream, and an occipital-parietal (i.e., dorsal) stream. Reciprocal connections exist between areas within these streams, and both pathways provide visual input to regions of prefrontal cortex that are involved in working memory and planning. At present, the functional significance of this organization for human behavior is not well-understood, and remains a source of active debate. Recent evidence from electrophysiology and neuroimaging point to the inadequacies of current theories, and suggest that areas within both streams may be considerably more interactive than previously supposed. Based on preliminary results, the present studies will evaluate an alternative model in which areas within the human dorsal and ventral streams are said to interact directly--via reciprocal interconnections, and/or indirectly-- through feedback from the prefrontal structures, during action planning.