Abstracts

Rationale: An increasing number of patients with intractable epilepsy are being treated with vagus nerve stimulator (VNS) and are still in pursuit of resective neurosurgery. Even if the stimulator is turned off for the duration of the MEG recording, unavoidable movement-related artifacts induced by VNS usually rendering the data worthless. In addition, significant artifact contamination can be caused due to prior craniotomies and dental work. In this study, we review the efficacy of the tSSS in severely contaminated MEG data of 24 patients with intractable epilepsy, referred between 2005 and 2008. Methods: MEG data were acquired with a 306-channel whole-head MEG system (ELEKTA-Neuromag, Finland). Simultaneous EEG was recorded using a 60-channel electrode array. Sleep was induced by prior sleep deprivation. Approximately 60 minutes of continuous data were recorded. The spatiotemporal signal space separation method (tSSS) [Taulu and Simola, 2006], an extended version of the spatial SSS [Taulu and Kajola, 2005], was used off-line to remove the artifacts. Results: Artifacts were caused due to implanted VNS devices in 11 patients, due to prior craniotomies in 6, due to dental work in 4, due to other sources in three. Spontaneous data quality obtained was excellent in 21 cases, satisfactory in two, and poor in one. Interpretation of the tSSS filtered data revealed abnormal findings in 19 cases, and was inconclusive in four. Conclusions: The tSSS filter is an important and effective tool for the analysis of artifact contaminated MEG data. This allows MEG scans for patients previously not considered suitable for MEG evaluations.