Abstract
Extrastriate area V4 neurons have receptive field (RF) profiles that are radially elongated. Response sensitivity gradients are steep on the side of the RF near the fovea and are stretched toward the periphery. Do the observed elongated response contours of V4 RFs represent a convergent sampling of the visual field that is different than the cortical magnification factor associated with area V1? To address this question V4 RFs were determined in 4 macaque monkeys. V4 RFs centered from 1.5 to 10 degrees in eccentricity were mapped with flashed stimuli using a 16 × 16 position grid adjusted for RF size. Two principle issues were addressed. 1) Do the V4 RFs represent a convergent weighted sampling of a circular patch of V1 relayed presumably through V2, and 2) are eccentricity dependent V4 RF size differences related to differences in the size of the sampled V1 patch? To visualize these issues the elongated V4 RF contours were projected onto a 3D model of the V1 surface. When projected onto the V1 surface the V4 RF contours appear as concentric circles. These observations were quantified by fitting the V4 receptive fields to an elliptical sampling model on the curved V1 surface. Receptive fields of area V4 neurons appear to represent a sampling of the V1 representation of the visual field corresponding to a circularly symmetric gaussian sampling with a SD of about 5 mm of cortical extent that is independent of eccentricity. Consequently except for a simple gain function, the cortical magnification factor for V4 neurons is the same as it is for V1 neurons, that is, there is no eccentricity dependent gain between V1 and V4.