EEG Classification by ICA Source Selection of Laplacian-Filtered Data

Claudio Carvalhaes, Marcos Perreau Guimaraes, Logan Grosenick, Patrick Suppes

Abstract: We studied the performance of a double-spatial filtering method for classification of single-trial electroencephalography (EEG) data that couples the spherical surface Laplacian (SL) and independent component analysis (ICA). This method was evaluated in the context of a binary classification experiment with brain states driven by mental imagery of auditory and visual stimuli. A statistically significant improvement was achieved with respect to the rates provided by raw data and by data filtered by either SL or ICA.