Evaluation of Signal Space Separation via simulation

Signal Space Separation (SSS) method is based on the solution of Laplace's equation: expansion of the spherical harmonics. It has been realized and proved that the series of vector spherical harmonics solution can be split into two parts that are related to the MEG (magnetoencephalography) measurements: brain signals and external interferences. In practice, SSS has become an elegant and promising method to recover bio-magnetic signal and remove external disturbance in empirical MEG measurements. In the present study, we evaluate SSS comprehensively via computer simulation and real applications. We generate two types of interference sources: magnetic dipole and electric current dipole. We examine the suppression effect for interference signals, and the goodness of the reconstruction for the interested signal. The expansion order of vector spherical harmonics plays critical role in both signal reconstruction in the brain and external artifact reduction, this is studied thoroughly and demonstrated in this paper. Besides computer simulation, SSS method is also applied on real MEG data sets and the effectiveness is investigated. Finally, we give an objective conclusion of the advantage and limitation of SSS method, and assess its value in practical MEG measurements.