Abstracts

Rationale: High frequency activity has been found to be a good predictor of the seizure onset zone. There is also a growing appreciation that seizures usually arise from epileptic networks, rather than a single focal source, and that the variability of seizures may be due to variations in the location of seizure onset and propagation within a larger network. We hypothesize that analysis of the variability of high frequency propagation during seizures can uncover the probabilistic dynamics of epileptic networks, and provide crucial information about the relative role of different cortical sites to these dynamics. Methods: Intracranial EEG recordings were obtained during pre-ictal and ictal intervals of multiple seizures (3-86) from six patients with intracranial electrodes. To investigate the propagation of high frequency activity (70-175 Hz) during these seizures, ictal signals were parameterized using multivariate autoregressive models (MVAR), and analyzed with a version of short-time direct Directed Transfer Function (SdDTF) computed from each ictal recording. SdDTF estimates the direction, intensity, and temporal evolution of the propagation among cortical sites. Each patient's ictal onset zone was identified independently by visual inspection of the ECoG recordings by experienced epileptologists. Results of SdDTF analyses were compared to the clinical evaluation. Results: Ictal patterns of high frequency propagation revealed variability among seizures. Some recording sites were involved during several seizures, but other sites were involved in only one seizure. In most seizures, the strongest propagation originated in sites identified by epileptologists as the ictal onset zone. However, we also observed ictal patterns in which strong propagation originated in sites not included in the seizure onset zone. These phenomena were observed in patients with spatially restricted ictal onsets, as well as those with widely distributed ictal onsets. Interestingly, we observed similar variability during pre-ictal intervals. Conclusions: Propagation of high frequency activity (70-175 Hz) may reveal variability in the sites of ictal origination. Varying patterns of ictal high-frequency propagation may be useful in estimating the ictal onset zone for epilepsy surgery. This project was funded by NINDS R01 NS40596.