Abstract
Synchrony has been studied with artificial stimuli (gratings, rings), but its behavior in the context of natural vision remains unknown. One challenge in analyzing responses to natural stimuli lies in linking response changes to a specific stimulus attribute. Spatial, temporal, and luminance properties may affect responses in many ways. To investigate the properties of natural stimuli that encourage assembly formation, we conducted differential measurements. Our general strategy was to measure a response to a control image and then modify the image by noise-addition and polarity-reversal. Cooperative responses were compared to those from the original image and the difference represented the consequences of the imposed modification. Using a 10x10 microelectrode array, we recorded from 93 groups of neurons (each containing 3–6 complex cells) in the visual cortex of paralyzed and anesthetized cats. Dilution of structure in the image with 50% added noise reduced synchrony by 33.7% (p < 0.004). Synchronous activity did not change for the polarity-reversed image (p > 0.05), as expected due to the polarity-invariance of complex cells. Average firing rates for each cell did not change significantly when compared to the original image for either the noise-added modification or the polarity-reversed modification (p > 0.05). Therefore, the degradation of structure by noise was represented more strongly by synchrony than by average firing rate. We have previously shown that synchrony between cell pairs can encode co-linear and co-circular contours and the current findings suggest that synchrony may be involved in signaling higher-order stimulus features like structure in natural scenes.
Supported by NIH: RO1 EY014680-03