Magnetoencephalography in partial epilepsy: clinical yield and localization accuracy.

The goals of this study were to determine (1) the yield of magnetoencephalography (MEG) according to epilepsy type, (2) if MEG spike sources colocalize with focal epileptogenic pathology, and (3) if MEG can identify the epileptogenic zone when scalp ictal electroencephalogram (EEG) or magnetic resonance imaging (MRI) fail to localize it. Twenty-two patients with mesial temporal (10 patients), neocortical temporal (3 patients), and extratemporal lobe epilepsy (9 patients) were studied. A 37-channel biomagnetometer was used for simultaneously recording MEG with EEG. During the typical 2-3-hour MEG recording session, interictal epileptiform activity was observed in 16 of 22 patients. MEG localization yield was greater in patients with neocortical epilepsy (92%) than in those with mesial temporal lobe epilepsy (50%). In 5 of 6 patients with focal epileptogenic pathology, MEG spike sources were colocalized with the lesions. In 11 of 12 patients with nonlocalizing (ambiguous abnormalities or normal) MRI, MEG spike sources were localized in the region of the epileptogenic zone as ultimately defined by all clinical and EEG information (including intracranial EEG). In conclusion, MEG can reliably localize sources of spike discharges in patients with temporal and extratemporal lobe epilepsy. MEG sometimes provides noninvasive localization data that are not otherwise available with MRI or conventional scalp ictal EEG.