Introduction: Magnetoencephalography (MEG) it is a non invasive technique of recording the brain activity, based on the detection of magnetic fields. Aim. It is sought to determine the relative contribution of MEG to the handling of patients with epilepsy concerning to other diagnostic technics like scalp EEG, v EEG (v EEG), electrocorticography (EcoG), magnetic resonance (MR and MRf), positron emission tomography (PET) and single photon emission tomography (SPECT).
Method: It has been carried out it a critical revision of the literature, using the approaches of the evidence based medicine (EBM). In the databases MEDLINE and HEALTH START, among the years 1996 and 2001, it were found 1169 references. After the one sieved, 36 were selected, of those that chose 22 definitively for their analysis: 3 articles were classified as of Technical Effectiveness and 19 as Effectiveness in Diagnostic Accuracy. For the methodological quality, 2 were classified as of group B, 16 in C and 4 in D. Eighteen articles make reference to MEG and pharmaco resistant epilepsy, while other 4 they make reference to MEG and pharmaco sensitive epilepsy. Sensibility is highly variable, being bigger in extratemporal epilepsy (44 92.9%), then in temporal (33.3 80%) and smaller in mesial temporal (42.3 53%). Specificity is bigger for patient with anatomical lesions.
Conclusion: 1. Localization of interictal activity is better in neocortical lateral regions, diminishing in mesial ones, 2. The methodological deficiencies of the articles prevent to carry out a meta analysis of the data.