Seventy patients were prospectively evaluated by simultaneously recorded MEG/EEG. All patients were surgical candidates or were considered for invasive EEG monitoring and had undergone an extensive presurgical evaluation at a tertiary epilepsy center. MEG and EEG raw traces were analysed individually by two independent reviewers.
MEG data could not be evaluated due to excessive magnetic artefacts in three patients (4 % ). In the remaining 67 patients, the overall sensitivity to detect IED was 72 % (48/67 patients) for MEG and 61 % for EEG (41/67 patients) analysing the raw data. In 13 % (9/67 patients), MEG-only IED were recorded, whereas in 3 % (2/67 patients) EEG-only IED were recorded. The combined sensitivity was 75 % (50/67 patients).
Three hundred and six-channel MEG has a similarly high sensitivity to record IED as EEG and appears to be complementary. In one-third of the EEG-negative patients, MEG can be expected to record IED, especially in the case of lateral neocortical epilepsy and/or cortical dysplasia.