Abstracts

Rationale: Interictal epileptiform discharges (IEDs), also known as interictal spikes, are intermittent, large-amplitude electrographic events recorded from epilepsy patients. Despite the relative prevalence of IEDs, compared to seizures, little is known about how IEDs relate to the seizure onset zone in human focal epilepsy. Two recent papers that leveraged microelectrode array recordings from human epilepsy patients independently measured consistent propagation directions of ictal discharges. Recent theoretical work predicted that as a consequence of such robust, spatially directed propagation of intense synaptic activity during seizures, IEDs from tissue surrounding the seizure onset zone learned to propagate in the same direction as ictal discharges. Critically, the geometry of focal seizure expansion dictates that direction points towards the center of the seizure onset zone. Methods: In order to test the hypothesis that IEDs are traveling waves, we employed a dataset of microelectrode array (96 electrodes in 4 x 4 mm, penetrating 1 mm) recordings from human subjects with medically refractory epilepsy. We detected IEDs from these microelectrode recordings by filtering the signal in the beta range and retaining excursions greater than 8 standard deviations that occurred across at least 10 microelectrodes within a window of 250 milliseconds. We then measured IED traveling wave speeds and directions using a previously validated multilinear regression framework. This framework fits a plane to the relative times of voltage maxima recorded across the microelectrode array during each IED. IEDs with model slopes that were significantly different from zero, based on a permutation test (500 permutations) were operationally defined as interictal traveling waves. Hypotheses that within-subject IED propagation directions were non-uniform, were tested with Hermans-Rasson tests of circular non-uniformity. Results: Across 6 subjects, we detected mean ± STD = 1749.4 ± 1019.8 IEDs. Of these detected IEDs, mean ± STD = 1194.8 ± 648.7 (68.3 %) were classified as interictal traveling waves (permutation tests, FDR-corrected p < 0.05). Distributions of IED traveling wave directions were significantly non-uniform (i.e. flat on a circular) in each subject (Hermans-Rasson tests, all p < 0.05), indicating that interictal traveling waves generally have consistent directions within individuals. However, several IED direction distributions exhibited secondary modes. Conclusions: These results show that IEDs are traveling waves with consistent directions that preliminary analyses suggest point toward the seizure onset zone, as predicted by the computational model. These traveling waves could therefore eventually be used to predict the location of seizure onset. One potential caveat of these results is that the microelectrode measures direction from a small patch of cortex, which may obscure the larger context of IED traveling wave direction. Future work will also examine patterns in IED traveling waves across larger scale clinical recordings, traveling wave speeds, and the significance of multimodal direction distributions. This work constitutes the first step towards leveraging features of relatively abundant IEDs for predicting spatial properties of relatively rare seizures, and may have quickly translatable implications for surgical treatment of medically refractory focal epilepsy. Funding: R21 NS113031-01 (Rolston)