Abstracts

Rationale: The epileptic brain is increasingly associated with pathological functional networks. Furthermore, the epileptogenic focus can be reliable identified by the presence of high frequency oscillations (HFOs) in the intracranial electroencephalogram (iEEG). In this study, we investigated whether HFOs spatially coincide with altered functional network characteristics derived from iEEG. Methods: Twelve patients with temporal lobe epilepsy (TLE) underwent bilateral depth electrode EEG recordings. We assessed epileptic network properties by computing phase lag index (PLI) as a measure for functional connectivity, and eigenvector centrality measure (ECM) as a measure for centrality of each electrode in the theta frequency band. We compared PLI and ECM between hemispheres and iEEG electrodes and correlated the left-right mirror electrodes HFO rate difference to PLI and ECM differences (Spearman's rho correlation). Results: A higher PLI was observed in left compared to the right hemisphere, regardless of lateralization of seizure onset zone. No differences were found between affected and non-affected hemisphere. Further analysis revealed a negative correlation between the difference in HFOs (fast ripples with a frequency of 250-500 Hz) and the difference in PLI (r = -0.8; p <0.01) and ECM (r = -0.8; p <0.01). Conclusions: Both PLI and ECM are decreased in electrodes with HFOs. These results revealed that the epileptogenic focus, identified by the presence of fast ripples, is associated with an altered functional network characteristics. High fast ripples rates are associated with low connectivity measures. Possibly, iEEG electrodes with HFOs are functionally isolated from their neighbouring channels.