Physiologically motivated multiplex Kuramoto model describes phase diagram of cortical activity

We derive a two-layer multiplex Kuramoto model from weakly coupled Wilson-Cowan oscillators on a cortical network with inhibitory synaptic time delays. Depending on the coupling strength and a phase shift parameter, related to cerebral blood flow and GABA concentration, respectively, we numerically identify three macroscopic phases: unsynchronized, synchronized, and chaotic dynamics. These correspond to physiological background-, epileptic seizure-, and resting-state cortical activity, respectively. We also observe frequency suppression at the transition from resting-state to seizure activity.