MEG in epilepsy and pre-surgical functional mapping

Abstract

Magnetoencephalography (MEG) is routinely used in pre-surgical evaluation of epilepsy. Magnetic source imaging (MSI) of epileptic spikes provides additional information to those provided by other non-invasive measures, including fluorodeoxyglucose-positron emission tomography (FDG-PET) and ictal single-photon emission computed tomography (SPECT), especially in neocortical epilepsy and in MRI-negative epilepsy. MSI may guide additional electrode coverage for intracranial EEG and area of resection when planning surgery; both of these approaches are associated with better seizure outcome. Mono-focal spike localization strongly indicates the epileptogenic zone. Complete removal of the MEG focus often results in the patient being seizure free, post-operatively. Similarities and differences between MEG and EEG should be well recognized when using MEG. Although the overall sensitivity of MEG to epileptic spikes is similar to that of EEG, such sensitivity can depend primarily on the orientation of equivalent current dipoles (ECD) of spikes. Favorable areas for MEG include the orbito-frontal, opercular, interhemispheric, temporo-lateral, and rolandic regions. MEG is less sensitive to deep regions, such as mesial temporal structures. MEG is also utilized for functional brain mapping. Somatosensory evoked fields to median nerve stimulation lead to an accurate, within a few millimeters, identification of the central sulcus. MEG analysis of event-related potentials or event-related de/synchronization in response to language tasks provides more than 80 % sensitivity and specificity in language lateralization for intra-carotid amobarbital procedures. MEG is a non-invasive alternative for pre-surgical determination of the language-dominant hemisphere.