A unified and quantitative network model for spatial attention in area V4

Abstract

Electrophysiological experiments in visual area V4 have shown that spatial attention induces a number of neural activity modulations. Depending on the stimulus characteristics, neuronal firing rates either increase or decrease. At the network level, the oscillatory activity in the gamma frequency range (30-70Hz) is enhanced by attention. Recently, pyramidal neurons and interneurons have been surmised to respond differently, but have been shown to have both a high firing variability. These results raise the question of the nature of the modulatory attentional input to V4 and of the network mechanisms that lead to the emergence of these different modulations. Here, we propose a biophysical network model of V4. We first reproduce the neural activity observed in response to different stimulus configurations. We found that different forms of the attentional input are possible, and that this fact could explain the observed multiplicity of modulations when stimulus contrast is varied. Our model offers a unified and quantitative picture, from which the cognitive roles played by these attentional modulations can be investigated.