Abstracts

Rationale: The earliest electrical evidence of seizure activity best represents the seizure onset zone. Accurate localization of a focal seizure onset zone provides the best prognosis after resective surgery. Very low and very high EEG frequencies may have an association with the initial ictal EEG changes recorded from intracranial electrodes. We retrospectively evaluated a patient who had undergone intracranial video/EEG monitoring employing wide-band frequency filters. We investigated the presumed epileptogenic zone determined by conventional recording settings, and compared with very low (ISP) and very high frequency (HFO) filter settings.Methods: We analyzed the initial ictal EEG changes in 7 habitual seizures with a sampling rate of 2000Hz, and a recording pass band of 0.016 to 600 Hz. The preictal period was evaluated for high and low frequency ictal changes from 2 minutes before the seizure onset time (as determined from the conventional settings). HFOs were identified based on a minimum of 100Hz frequency and at least 0.1 second duration. ISPs were defined as positive or negative shifts of at least 300uV in amplitude, lasting at least 3 seconds, and consistently occurring in the same location within +/- 30 seconds of the initial ictal EEG changes. Ictal EEG onset time and electrode localizations based on the agreement between two independent reviewers were also determined. Results: This is 22-year old female with refractory epilepsy who has had seizures since age of 10 and had 3 prior right temporal lobe resections (2000, 2001, and 2004) with no specific pathology. Her seizures are consisted of epigastric aura following autonomic seizures with hypersalivation but preserved awareness. MRI (2008) revealed hyperintense soft tissue remnants of the superior temporal gyrus and the hippocampal head and body. Ictal SPECT/SISCOM (2008) highlighted areas in the right posterior basal medial temporal and occipital junction. A posterior mesial temporo-occipital origin was suspected, although the seizure semiology is more consistent with an insular onset. Intracranial video/EEG monitoring was performed with extensive coverage of right insula, basal temporo-occipital junction and mesial temporo-occipital regions with grid and depth electrodes. In all 7 seizures recorded with wide-band frequency filter, HFOs and ISPs were strikingly localized to the same or neighboring electrodes but different from the onset zone determined from the conventional filtering. (Figure A, B). Resective surgery was performed based on the conventional ictal EEG recordings, but two years later the patient continues to have seizures. Her postoperative MRI showed that the resection site did not include the area of ictal zone identified by the wide-band frequency filter recordings. Conclusions: Our study shows that ISPs may provide more useful information than conventional EEG settings without the difficulties of analyzing HFOs. Evaluation of ISPs may help to more conveniently localize the epileptogenic zone in refractory epilepsy.