Processing of kinetically defined boundaries in the cortical motion area MT of the macaque monkey.

Abstract

1. Electrophysiological recordings of 68 cells in the middle temporal area MT were made in paralyzed and anesthetized macaque monkeys. 2. Testing with our kinetic boundary stimuli always occurred under optimized conditions. To this end, the preferred direction, speed, stimulus position, and stimulus size of each cell were determined by quantitative tests. 3. The orientation selectivity to stationary luminance contrast edges served as a reference by which a response to kinetic boundaries could be compared. We found cells in area MT to be less selective to the orientation of luminance contrast stimuli than to the direction of motion. We confirmed the presence of neurons with preferred orientation aligned with their preferred direction. 4. The responses to kinetic edges defined by motion vectors moving in opposite directions, kinetic gratings with motion vectors in opposite directions, kinetic edges containing coherent motion and a stationary complementary field or coherent motion and a complementary field containing visual dynamic noise were compared. Kinetic boundaries were generated so that the motion vectors moved either parallel or orthogonal to the orientation of the discontinuity. For a cell to be considered as responding to the orientation of a kinetic boundary, it had to exhibit the same preferred orientation when the local motion vectors changed from parallel to orthogonal to the orientation of the kinetic boundary. 5. All cells in area MT changed their preferred orientation by 90 degrees when the coherent motion vectors changed from moving parallel to moving orthogonal to the boundary. This was the case independent of the types of kinetic boundary tested. We concluded that cells in area MT appear to respond to the motion vector over their classical receptive field (CRF) only and were unable to code the orientation of the kinetic boundary. 6. In those cells exhibiting an antagonistic surround, we examined the ability of the cell to code the position of a kinetic boundary. None of the cells tested signaled the position of a kinetic boundary. The side preference of the stimulus of the cells changed from left to right as the motion vectors in the stimulus reversed. This indicates that the cells were only selective for the motion vectors present over their CRF. 7. We found that the directional sensitivity of cells in area MT remained unaltered by the presence of additional motion vectors within the CRF. This suggests that cells in area MT extract a specific motion vector from a spatial configuration of vectors.(ABSTRACT TRUNCATED AT 400 WORDS)