Abstract
The responses of neurons in area V1 depend on stimulus attributes such as orientation and position. There is interest in knowing how these attributes combine to determine a response in populations of neurons. We asked whether the population responses to a localized oriented stimulus could be determined by independent measures of retinotopy and orientation selectivity. We measured population responses by imaging voltage-sensitive dye fluorescence in area V1 of anesthetized cats. Stimuli were contrast-reversing square wave gratings windowed in narrow rectangles elongated parallel to the axis of grating orientation. These stimuli activated regions of cortex that are broad but patchy. The center of the activity depended on the position of the stimulus and the patchiness resulted from the orientation of the stimulus. The width of the activated region, in turn, could be attributed to the point spread function of the cortex (i.e., the cortical representation of a point on the retina), which we modeled as a Gaussian (s.d. = 1.1+−0.4 mm). The population responses were well fit (∼74% of the variance) by a model that assumes multiplicative contributions of position and orientation preference, indicating that stimulus orientation and position contribute independently to the population responses in primary visual cortex.