Recently a major pathway for the analysis of visual motion has been identified in the visual cortex of monkeys. this pathway begins in layers 4b and 6 of striate cortex (V1), projects to the middle temporal area (MT), and then projects from MT to the medial superior temporal area (MST). In the experiments outlined in this proposal we will examine the processing of visual motion that takes place at the three cortical areas in this pathway.
The first aim will be to study direction-opponency in areas V1 and MT. (a) Transparent motion stimuli will be used as a tool to test a two stage model in which directions of motion are determined in V1, and cells tuned to different directions of motion inhibit one another in MT. A variety of two surface stimuli will be developed; some produce robust percepts of transparent, moving surfaces and other produce little or no percept of motion or transparency. The hypothesis will be tested that the stimuli that appear non-transparent produce maximum suppression at the opponent stage in MT, and as a result lead to a loss of perceived motion. (b) It will be determined whether MT neurons with similar spatial frequency and disparity selectivities are more likely to inhibit one another than cells with dissimilar selectivities. If so, one likely role of the opponent stage in MT is to reduce noise when perceiving motion, since opposed motion signals at the same depth or spatial frequency are likely to result from flicker (noise), whereas signals at different depths or spatial frequencies are likely to represent different surface in motion.
A second aim will be to determine whether area MT cells convey information about speed gradients. Such information is sufficient for perceiving the three-dimensional shapes of objects in motion. Lesions to area MT produce deficits in perceiving 3D shape from speed gradients, suggesting that it is a likely site for encoding this information. It will be determined whether MT neurons are selective for speed gradients, or whether they respond best to flat, translational motions.
A third aim i s to examine motion-pattern selectivity of area MST neurons to complex motion patterns. (a) Previous studies have indicated that MST neurons respond to expansion/contraction, rotation and translation motions. Experiments will test whether area MST decomposes the optical flow into three channels corresponding to these three motions, or if alternatively MST cells are tuned to a much wider variety of motion patterns. The pattern-selective properties of these cells will be further tested for position, speed, and size invariance. (b) A columnar organization for motion-pattern selectivity has been reported in area MST based on single cell recordings. the functional topography of area MST will be determined using a double-label 2-deoxyglucose technique. (c) The role of MST in perceiving motion patterns will be directly tested by examining whether microstimulation of these columns can bias perceptual discriminations of different motion patterns.
Showing the most recent 10 out of 40 publications
