Abstracts

Rationale: There is a continuing debate in the literature about existence of higher or lower spike rates before or after epileptic seizures. The underpinning of our study was the formulation of a mathematical framework within which spatio-temporal dynamics of epileptic spikes help elucidate clinically relevant questions. In particular, within this framework, we investigated the role of epileptic spikes in ictogenesis and the possibility to reliably localize the epileptogenic focus from interictal spiking in patients with focal epilepsy. Methods: Long-term (5-10 days) intracranial electroencephalograms (iEEGs) from 5 patients with temporal lobe epilepsy (TLE) were analyzed in this study. Epileptic spikes were detected from the iEEGs using an in-house developed Data Adaptive Morphological Filter (DAMF) algorithm. A novel spatial synchronization measure (SSM) of spikes was then developed and applied to iEEGs. (Both DAMF and SSM algorithms were first tested on simulation data from coupled spiking neuron models.) Localization of the epileptogenic focus from interictal periods was performed on the basis of SSM and measures from graph theory (eigenvector centrality) [1]. Results: Application of SSM to interictal iEEG revealed monotonically increasing long-term (order of hours) synchronization between spike trains at critical brain sites that included the focal zone. After seizures' end, spike synchronization decreased. This behavior is in agreement with our previously postulated hypothesis that seizures occur to reset a pathologically established preictal hypersynchrony of EEG dynamics between critical brain sites [2, 3]. It is to be noted that such findings were not statistically significant when spike rates instead of individual spikes were used for analysis. With respect to focus localization, the application of DAMF to interictal spiking activity resulted to accurate localization of the focus (p<0.05) in 4 out of our 5 patients. In the fifth patient, focus localization was not statistically significant (p>0.05) [1]. Conclusions: This study generated results that, if validated in a larger cohort of patients with focal epilepsy, could have immediate clinical applications. Preictal synchronization of individual spikes between critical brain sites and their postictal desynchronization were consistent findings across patients and seizures, as long as spiking activity could be detected in the EEG. These findings definitely render a physiological dimension to the theory of resetting of synchronized global brain dynamics at seizures [2]. Finally, the ability to localize the focus from spatio-temporal analysis of interictal spiking supports the notion that epileptogenic focus could be detected as the central node of an epileptic network at work even in interictal periods. [1] B. Krishnan, Ph.D. Dissertation, Arizona State University, May 2012 [2] S. Sabesan et al., J. Combinatorial Optimization, 17, pp. 74-97, 2009. [3] L. Iasemidis et al., IEEE Transactions on Biomedical Engineering, 51, pp. 493-506, 2004.