The Application of Spatiotemporal Energy Model in the Simulation of Population Responses in Early Visual Cortices

The early visual cortices (V1 and V2) are traditionally regarded as local feature detectors or filters of spatiotemporal components while higher visual areas as centers for global feature or object recognition in high mammalians. Natural scenes are complicated, thus synthesized visual stimuli are often composed of various local and global visual cues. Here we propose that the spatiotemporal energy model could simulate the population responses to the local component of most visual stimulus in the early visual cortices. The population responses to stimuli of either luminance or contrast and texture modulations, recorded using intrinsic optical imaging in both V1 and V2 of macaques and cats, could be successfully predicted by the energy model. However, it failed to predict the population responses to other complex stimuli such as illusory and kinetic contours. These results illustrated the application and limitation of the spatiotemporal energy model in accounting for population responses in the early visual cortices.