Abstracts

Rationale: Pathological high frequency oscillations (HFOs) have emerged as a biomarker for the identification of the epileptogenic zone (EZ). Detection of HFOs may improve presurgical diagnosis and surgical outcome of epilepsy patients. HFOs identify the seizure onset zone (SOZ) with higher sensitivity and specificity compared to the irritative zone, while the surgical removal of the HFOs-generating tissue correlates with better outcomes than the removal of the SOZ or the irritative zone. Here, we present evidence of spatiotemporal propagation of interictal HFOs in the range of ripples (80-250 Hz) detected with extraoperative electrocorticography (ECoG) from pediatric epilepsy surgery patients.Methods: Intracranial extraoperative recordings were obtained from 21 children (14 males and 7 females; age: 12.03 years ± 5.73) with medically resistant focal onset seizures as part of epilepsy surgery planning. Patients were selected with a broad spectrum of etiologies, EEG findings, and seizure patterns (see Table 1). All patients were seizure free one year after their surgery. Intracranial EEG was analyzed from baseline interictal recordings, selected from the first 50 minutes of extraoperative ECoG that was free from technical disruptions or clinical events. The multichannel data (total duration: 17 hrs 16 min) were filtered between 80 and 250 Hz by using a 3rd order Butterworth filter. An automated algorithm that extracts the HFOs from the envelope of the ECoG signal was used to detect the HFOs. Clusters of HFO events across the multichannel ECoG traces were subsequently analyzed for all patients in order to investigate their relative time delays and to infer their propagation. The HFO propagation was displayed on patients’ anatomical scans. The recording site bearing the smallest rank was labeled as the lead generator of HFO discharges. The HFOs were correlated with the SOZ as this was defined by the epileptologists.Results: We identified HFO bursts in 14 patients (see Table 1) with a mean frequency of 106.84 ± 11.66 Hz (Figure 1a and c). Visual inspection of the filtered data indicated a clear propagation pattern across channels in ten patients (see Figure 1a). The lead generators’ area of the HFOs propagation as it was defined by our algorithm (ECoG electrodes with low rank) was overlapping with the SOZ that was resected in eight patients (p<0.05) (see Figure 1b).Conclusions: We report a spatiotemporal propagation pattern of interictal HFOs in the ripple frequency range. The temporal lead generators of this activity overlap well with the SOZ as this was defined by the epileptologists. Our results indicate that the lead generators of HFOs may serve as a good presurgical biomarker that identifies reliably the EZ not only at seizure onset but also during the interictal period. Such a biomarker could augment the results of conventional long term monitoring and lead to a significant improvement of the presurgical evaluation procedure. Source of Funding: This study is supported by the Epilepsy Foundation / American Epilepsy Society Research Grant.