Added diagnostic value of magnetoencephalography (MEG) in patients suspected for epilepsy, where previous, extensive EEG workup was unrevealing

Objective: To elucidate the possible additional diagnostic yield of MEG in the workup of patients with suspected epilepsy, where repeated EEGs, including sleep-recordings failed to identify abnormalities.

Methods: Fifty-two consecutive patients with clinical suspicion of epilepsy and at least three normal EEGs, including sleep-EEG, were prospectively analyzed. The reference standard was inferred from the diagnosis obtained from the medical charts, after at least one-year follow-up. MEG (306-channel, whole-head) and simultaneous EEG (MEG-EEG) was recorded for one hour. The added sensitivity of MEG was calculated from the cases where abnormalities were seen in MEG but not EEG.

Results: Twenty-two patients had the diagnosis epilepsy according to the reference standard. MEG-EEG detected abnormalities, and supported the diagnosis in nine of the 22 patients with the diagnosis epilepsy at one-year follow-up. Sensitivity of MEG-EEG was 41%. The added sensitivity of MEG was 18%. MEG-EEG was normal in 28 of the 30 patients categorized as 'not epilepsy' at one year follow-up, yielding a specificity of 93%.

Conclusions: MEG provides additional diagnostic information in patients suspected for epilepsy, where repeated EEG recordings fail to demonstrate abnormality.

Significance: MEG should be included in the diagnostic workup of patients where the conventional, widely available methods are unrevealing.