Modulation of sensory suppression: Implications for receptive field sizes in the human visual cortex

Neurophysiological studies in monkeys show that when multiple visual stimul i appear simultaneously in the visual field, they are not processed indepen dently, but rather interact in a mutually suppressive way. This suggests th at multiple stimuli compete for neural representation. Consistent with this notion, we have previously found in humans that functional magnetic resona nce imaging (fMRI) signals in V1 and ventral extrastriate areas V2, V4, and TEO are smaller for simultaneously presented (i.e., competing) stimuli tha n for the same stimuli presented sequentially (i.e., not competing). Here w e report that suppressive interactions between stimuli are also present in dorsal extrastriate areas V3A and MT, and we compare these interactions to those in areas V1 through TEO. To exclude the possibility that the differen ces in responses to simultaneously and sequentially presented stimuli were due to differences in the number of transient onsets, we tested for suppres sive interactions in area V4, in an experiment that held constant the numbe r of transient onsets. We found that the fMRI response to a stimulus in the upper visual field was suppressed by the presence of nearby stimuli in the lower visual field. Further, we excluded the possibility that the greater fMRI responses to sequential compared with simultaneous presentations were due to exogeneous attentional. cueing by having our subjects count T's or L 's at fixation, an attentionally demanding task. Behavioral testing demonst rated that neither condition interfered with performance of the T/L task. O ur previous findings suggested that suppressive interactions among nearby s timuli in areas V1 through TEO were scaled to the receptive field (RF) size s of neurons in those areas. Here we tested this idea by parametrically var ying the spatial separation among stimuli in the display. Display sizes ran ged from 2 x 2 degrees to 7 X 7 degrees and were centered at 5.5 degrees ec centricity. Based on the effects of display size on the magnitude of suppre ssive interactions, we estimated that RF sizes at an eccentricity of 5.5 de grees were <2<degrees> in V1, 2-4 degrees in V2, 4-6 degrees in V4, larger than 7 degrees (but still confined to a quadrant) in TEO, and larger than 6 degrees (confined to a quadrant) in V3A. These estimates of RF sizes in hu man visual cortex are strikingly similar to those measured in physiological mapping studies in the homologous visual areas in monkeys.