Use of common average reference and large-Laplacian spatial-filters enhances EEG signal-to-noise ratios in intrinsic sensorimotor activity

Abstract Background Oscillations in the resting-state scalp electroencephalogram (EEG) represent various intrinsic brain activities. One of characteristic EEG oscillations is sensorimotor rhythm (SMR)—with its arch-shaped waveform alpha- and betabands—that reflect activity. The representation activity by SMR depends on signal-to-noise ratio achieved spatial filters. New method We employed simultaneous recording functional magnetic resonance imaging, 10-min activities were recorded 19 healthy volunteers. To compare spatial-filtering methods commonly used for extracting cortical activities, we assessed nine different spatial-filters: a default reference amplifier system, common average (CAR), small-, middle- large-Laplacian filters, four types bipolar manners (C3-Cz, C3-F3, C3-P3, C3-T7). identified region that correlated with EEG-SMR power obtained after each was applied. Subsequently, calculated proportion significant voxels cortex as well occupancy all regions to examine sensitivity specificity spatial-filter. Results CAR spatial-filters superior at improving ratios from signal. Comparison existing Our results are consistent spatial-filter selection extract task-dependent activation better control EEG-SMR-based interventions. approach has potential identify optimal EEG-SMR. Conclusions Evaluating spontaneous useful procedure constructing more effective