A model of direction selectivity in cortical simple cells based on lagged thalamic input and intracortical feedback

A model of a single hypercolumn in the primary visual cortex for the analysis of simple cell orientation and direction selectivity is proposed. The model embodies a feedforward mechanism, based on the convergence of lagged and nonlagged inputs onto a cortical cell, to reproduce the spatial and temporal structure of the receptive field, and lateral (inhibitory and excitatory) connections within the hypercolumn. It explains both orientation and direction selectivity of simple cells. The response properties of the two types of inputs are described by the product of linear spatial and temporal response functions. To characterize the temporal response properties of lagged and nonlagged neurons we have used pure time delays. The lateral connections are arranged in a push-pull fashion. Simulation results show that the model is able to mimic several experimental data. In particular, it is able to reproduce the response of simple cortical cells, with space-time inseparable receptive field, characterized by a strong direction selectivity.