Neural Mechanisms of Selective Visual Attention

Massachusetts Institute Of Technology, Cambridge MA

Investigators

Linked publications & trials

Abstract

[unreadable] DESCRIPTION (provided by applicant): When people are confronted by a typical, crowded visual scene, attentional mechanisms are needed to limit visual processing to objects that are currently relevant to behavior. An understanding of these attentional mechanisms will ultimately help in developing a visual prosthesis for people with severe visual impairments, and will also help in developing treatments for people with brain disorders affecting attention, including ADHD. The long term goal of this proposal is to not only uncover attentional influences on visual processing in the brain but also to understand the neuronal mechanisms by which these effects occur. Our focus is on visual areas V1, V2, and V4, which are among the earliest processing stages in the cortical pathway important for object recognition. New results suggest that the attentional bias in favor of behaviorally relevant stimuli may involve not only changes in the average firing rate of neurons but also the timing of neural activity in different cortical layers. The first specific aim is to identify the effects of spatially-directed attention on the timing of spikes in populations of cells in V4, and relate these effects to specific visually- guided behaviors. The next aim is to differentiate the effects of spatially directed attention on cells in the superficial versus deep lamina of areas V1, V2, and V4. The third aim is to establish the sequence of the effects of spatially directed attention on the visual pathway extending from V1 through area V4. The central hypotheses of the research program are that top-down attentional mechanisms induce changes in not only the firing rate but also the temporal synchrony of responses in extrastriate cortex, that these attentional effects on neural activity have functional correlates, and that these changes are relayed in a "backward" direction from higher order areas to lower order ones. These hypotheses test novel ideas, but they are well supported by preliminary data, and we are well-positioned to test them since we have had many years of experience with the techniques. [unreadable] [unreadable] [unreadable]