External Seminar: James Rankin

[Centre for Mathematical Medicine and Biology Seminar]

James Rankin

Pattern formation in a neural field model of visual cortex

We study localised states in the neural field equation (NFE) posed on the Euclidean plane. The resulting nonlocal integro-differential equation has been widely used to model the mean firing rates of neurons across a spatially continuous domain. Primary visual cortex features a quasiperiodic orientation preference map with a regular length scale. Voltage imaging experiments have shown that local, oriented visual stimuli elicit activation that is orientation-selective and patchy within the stimulus footprint but non-selective outside the footprint. We study the dynamics of these input-driven states in a model with a biologically-motivated radial connectivity profile and sub-populations encoding different orientations. We argue that the observed patchy cortical activation patterns are pre-encoded by the connectivity profile and that the orientation preference map fixes the spatial phase of such patterns.