Evaluation of signal space separation via simulation.

Signal space separation (SSS) method is an advanced signal-processing approach that can be used to recover bio-magnetic signal and remove external disturbance in empirical magnetoencephalography (MEG) measurements. SSS is based on the solution of the quasi-static approximation of Maxwell equations (i.e., Laplace's equation) which can be expressed as linear combinations of spherical harmonic functions. In applying SSS, MEG measurements can be split into two parts: brain signals and external interferences. In this paper, after a brief review of the basics of SSS, we evaluate SSS systematically via computer simulation and real MEG data. In the simulations of this paper, two types of interference sources with magnetic and electric current dipoles are used. The interference suppression effects and the quality of the reconstruction of the interested signal are investigated. Also, the degree of spherical harmonic functions and its relationship with signal reconstruction and interference suppression are studied thoroughly. Finally, we provide objective assessments of the advantages and limitations of the SSS approach, and its practical value in MEG measurements.