Abstracts

Rationale: A growing body of evidence supports a model of epilepsy as a network disorder. From this perspective, seizure development is driven by large-scale network dysfunction rather than the activity of a single epileptic focus. As such, regions outside of the seizure focus may play pivotal roles in early seizure development and spread. Abnormal interactions between the seizure focus and outside regions during the transition between seizure (ictal) and non-seizure (interictal) states may serve as effective targets for neuromodulation designed to disrupt seizure development. It has been shown that coherence during ictal states is different from that of inter-ictal states. However, how this change from interictal to ictal occurs is not well understood. During this ‘preictal’ transition to seizure, brain regions outside of the seizure focus are recruited into synchronous activity. We ask whether this synchronization results from increased influence from the focus to sites outside the focus, or from outside sites to the seizure focus.

Methods: We analyze intracranial EEG (iEEG) data from patients undergoing phase II monitoring for future surgical treatment of pharmaco-resistant epilepsy. Continuous iEEG data issegmented based upon temporal proximity to seizure. For each channel, we examine two-minute preictal segments immediately preceding the channel’s recruitment into synchronous activity. Granger Causality Index (GCI), a time-varying measure of directed functional connectivity, is calculated across duration ofthe pre-ictal segment. Two sets of time courses are generated: GCI time courses representing the influence of the activity from each channel outside the focus on the activity of each channel within the focus and those representing the influence of the activity within the focus on activity recorded at outside sites.The ratio between GCI from the seizure focus to outside sites relative to GCI in the opposite direction was tracked throughout the preictal period.

Results: Our preliminary results suggest that, on average, the ratio between GCI from the seizure focus to outside sites relative to GCI in the opposite direction is approximately 1.2 times larger during the interictal period relative to preictal periods.

Conclusions: Initial results support that influence from the seizure focus to outside sites may paradoxically decrease immediately prior to a channel’s recruitment into synchronous seizure activity. Alternatively, this effect may be mediated by increasing influence from outside sites to the seizure focus during seizure development. This research lays groundwork for development of network-informed neuromodulation for epilepsy. The results suggest that neuromodulation disrupting connectivity between epileptic foci and downstream sites may serve as optimal targets for early disruption of seizure development.

Funding: Floyd and Mary Schwall Fellowship for Medically Related Research