Evaluation of a dual signal subspace projection algorithm in magnetoencephalographic recordings from patients with intractable epilepsy and vagus nerve stimulators

ABSTRACT Magnetoencephalography (MEG) data is subject to many sources of environmental noise, and interference rejection is a necessary step in the processing of MEG data. Large amplitude interference caused by sources near the brain have been common in clinical settings and are difficult to reject. Artifact from vagal nerve stimulators (VNS) is a prototypical example. In this study, we describe a novel MEG interference rejection algorithm called dual signal subspace projection (DSSP), and evaluate its performance in clinical MEG data from people with epilepsy and implanted VNS. The performance of DSSP was evaluated in a retrospective cohort study of patients with epilepsy and VNS who had MEG scans for source localization of interictal epileptiform discharges. DSSP was applied to the MEG data and compared with benchmark for performance. We evaluated the clinical impact of interference rejection based on human expert detection and estimation of the location and time‐course of interictal spikes, using an empirical Bayesian source reconstruction algorithm (Champagne). Clinical recordings, after DSSP processing, became more readable and a greater number of interictal epileptic spikes could be clearly identified. Source localization results of interictal spikes also significantly improved from those achieved before DSSP processing, including meaningful estimates of activity time courses. Therefore, DSSP is a valuable novel interference rejection algorithm that can be successfully deployed for the removal of strong artifacts and interferences in MEG. HIGHLIGHTSPerformance evaluation of a novel MEG interference rejection algorithm called dual signal subspace projection (DSSP) in clinical MEG data from patients with epilepsy and implanted vagus nerve stimulators (VNS).DSSP shows significant low‐frequency reduction of artifacts from VNS, and increases the detection and identification of interictal epileptic spikes.DSSP improved the timing and localization of identified epileptic spikes.DSSP is a valuable novel interference rejection algorithm that can be successfully deployed for the removal of strong artifacts and interferences in MEG.

[1]  Hagai Attias,et al.  Partitioned Factor Analysis for Interference Suppression and Source Extraction , 2006, ICA.

[2]  Y. Adachi,et al.  Reduction of non-periodic environmental magnetic noise in MEG measurement by continuously adjusted least squares method , 2001 .

[3]  S. Taulu,et al.  Spatiotemporal signal space separation method for rejecting nearby interference in MEG measurements , 2006, Physics in medicine and biology.

[4]  B. Widrow,et al.  Adaptive noise cancelling: Principles and applications , 1975 .

[5]  Lucas C. Parra,et al.  Joint decorrelation, a versatile tool for multichannel data analysis , 2014, NeuroImage.

[6]  Kensuke Sekihara,et al.  A probabilistic algorithm for robust interference suppression in bioelectromagnetic sensor data , 2007, Statistics in medicine.

[7]  David P. Wipf,et al.  A unified Bayesian framework for MEG/EEG source imaging , 2009, NeuroImage.

[8]  Hagai Attias,et al.  Probabilistic algorithms for MEG/EEG source reconstruction using temporal basis functions learned from data , 2008, NeuroImage.

[9]  Kensuke Sekihara,et al.  Subspace-based interference removal methods for a multichannel biomagnetic sensor array , 2017, Journal of neural engineering.

[10]  Julia P. Owen,et al.  Performance evaluation of the Champagne source reconstruction algorithm on simulated and real M/EEG data , 2012, NeuroImage.

[11]  G. Deuschl,et al.  Recommendations for the practice of clinical neurophysiology: guidelines of the International Federation of Clinical Neurophysiology. , 1999, Electroencephalography and clinical neurophysiology. Supplement.

[12]  Kensuke Sekihara,et al.  Dual signal subspace projection (DSSP): a novel algorithm for removing large interference in biomagnetic measurements , 2016, Journal of neural engineering.

[13]  John S Ebersole,et al.  American Clinical Magnetoencephalography Society Clinical Practice Guideline 1: Recording and Analysis of Spontaneous Cerebral Activity* , 2011, Journal of clinical neurophysiology : official publication of the American Electroencephalographic Society.

[14]  Julia P. Owen,et al.  Robust Bayesian estimation of the location, orientation, and time course of multiple correlated neural sources using MEG , 2010, NeuroImage.

[15]  Hagai Attias,et al.  A graphical model for estimating stimulus-evoked brain responses from magnetoencephalography data with large background brain activity , 2006, NeuroImage.

[16]  S. Baillet,et al.  Automated interictal spike detection and source localization in magnetoencephalography using independent components analysis and spatio-temporal clustering , 2004, Clinical Neurophysiology.

[17]  Riitta Hari,et al.  Removal of magnetoencephalographic artifacts with temporal signal‐space separation: Demonstration with single‐trial auditory‐evoked responses , 2009, Human brain mapping.