Validation of head movement correction and spatiotemporal signal space separation in magnetoencephalography

Objective: Our aim was to assess the effectiveness and reliability of spatiotemporal signal space separation (tSSS) and movement correction (MC) in magnetoencephalography (MEG) recordings disturbed by head movements and magnetized material on the head.

Methods: We recorded MEG from 20 healthy adults in stationary (reference) head position and during controlled head movements. Nearby magnetic interference sources were simulated by attaching magnetized particles on the subject's head. Auditory and somatosensory stimuli were presented. MC, tSSS and averaging were performed to obtain auditory (AEF) and somatosensory (SEF) evoked fields. Neuronal sources were modeled as equivalent current dipoles. MC was also validated by reconstructing signals generated by current dipoles in a phantom.

Results: After MC, the AEF and SEF responses recorded during intermittent head movements were similar in amplitude to the reference recordings and differed by 5-7mm in source location. The tSSS method removed artifacts due to the attached magnetized particles but did not affect the reference data.

Conclusions: The methods are able to reliably recover MEG responses contaminated by movements and magnetic artifacts on the head.

Significance: The combination of tSSS and MC methods is especially useful in clinical measurements, where movements and magnetic disturbances are commonly present.