Abstracts

Rationale: To compare the results of determination of seizure onsets estimated from source analysis using ictal recordings from noninvasive 256 channel, dense array long-term EEG-video monitoring (LTM) to that obtained from subsequent intracranial LTM. Methods: Nine patients with medically refractory localization-related epilepsy, all surgical candidates, underwent intracranial LTM (subdural grid or strip electrodes or both) after standard LTM failed to provide adequate localizing information regarding ictal origins. Prior to invasive studies all subjects underwent dense array LTM (48-96 hrs). Habitual clinical seizures were recorded in all cases during dense array LTM; source analysis was applied to ictal onsets and the results were displayed on standard MRI. Patient mean age was 23 yrs (range 10-49); seven were males. Known risk factors included a history of closed head injury in one case, and childhood meningitis in another. No focal neurological signs were found in any patient. MRI was normal in six; one had cerebellar hypoplasia, one had right frontal-parietal dyplasia, and another, multiple bilateral cavernous angiomas. Results: Ictal onsets, based on invasive recordings, were of medial temporal lobe onset in three patients, lateral parietal cortical origin in two, medial parietal in one, lateral frontal in one, frontopolar in one, and posterior inferior temporal-occipital onset in one. Dense array EEG findings revealed convergent localization of ictal onsets in 7/9 cases and similar lateralization in all patients. Surgical resections were based on the invasive EEG findings. Conclusions: Dense array EEG has the potential to assist in the noninvasive localization of epileptic seizures, and in guiding the placement of intracranial EEG electrodes, when standard methods of evaluation fail to provide adequate information on ictal origin.