Comparison of DSSP and tSSS algorithms for removing artifacts from vagus nerve stimulators in magnetoencephalography data.

Abstract

OBJECTIVE\nLarge amplitude artifacts from vagus nerve stimulator (VNS) implants for refractory epilepsy affect magnetoencephalography (MEG) recordings and are difficult to reject, resulting in unusable data from this important population of patients who are frequently evaluated for surgical treatment of epilepsy. Here we compare the performance of two artifact removal algorithms for MEG data: the dual signal subspace projection (DSSP) and temporally extended signal space separation method (tSSS) algorithms.\n\n\nAPPROACH\nEach algorithm s performance was first eval- uated in simulations. We then tested the performance of each algorithm on resting-state MEG data from patients with VNS implants. We also examined how each algorithm improved source localization of somatosensory evoked fields (SEFs) in patients with VNS implants.\n\n\nMAIN RESULTS\nDSSP and tSSS algorithms have similar ability to reject interference both in simulated and in real MEG data, if the origin location for tSSS is appropriately set. If the origin set for tSSS is inappropriate, the signal after tSSS can be distorted, due to a mismatch between the internal region and the actual source space. Both DSSP and tSSS are able to remove large amplitude artifacts from outside the brain. DSSP might be a better choice than tSSS when choice of origin location for tSSS is difficult.\n\n\nSIGNIFICANCE\nBoth DSSP and tSSS algorithms can recover distorted MEG recordings from people with intractable epilepsy and VNS implants, improving epileptic spike identification and source localization of both functional activity and epileptiform activity. .