Predicting the spatiotemporal diversity of seizure propagation and termination in human focal epilepsy

Abstract: Recent studies have shown that seizures can spread and terminate across brain areas via a rich diversity of spatiotemporal patterns. In particular, while the location of the seizure onset area is usually in-variant across seizures in a same patient, the source of traveling (2-3 Hz) spike-and-wave discharges (SWDs) during seizures can either move with the slower propagating ictal wavefront or remain stationary at the seizure onset area. In addition, although most focal seizures terminate quasi-synchronously across brain areas, some evolve into distinct ictal clusters and terminate asynchronously. To provide a unifying perspective on the observed diversity of spatiotemporal dynamics for seizure spread and termination, we introduce here the Epileptor neural field model. Two mechanisms play an essential role. First, while the slow ictal wavefront propagates as a front in excitable neural media, the faster SWDs propagation results from coupled-oscillator dynamics. Second, multiple time scales interact during seizure spread, allowing for low-voltage fast-activity (>10 Hz) to hamper seizure spread and for SWD propagation to affect the way a seizure terminates. These dynamics, together with variations in short and long-range connectivity strength, play a central role on seizure spread, maintenance and termination. We demonstrate how Epileptor field models incorporating the above mechanisms predict the previously reported diversity in seizure spread patterns. Furthermore, we confirm the predictions for synchronous or asynchronous (clustered) seizure termination in human seizures recorded via stereotactic EEG. Our new insights into seizure spatiotemporal dynamics may also contribute to the development of new closed-loop neuromodulation therapies for focal epilepsy.