Clinical applications of the signal space separation method

Abstract

Conventional biomagnetic measurements suffer from the extreme sensitivity of the method to external and subject-related interference. Especially difficult to remove artifacts are caused by magnetic sources inside the shielded room and especially by magnetization associated with the subject. Such artifacts are encountered, e.g. with subjects having braces, magnetic stimulators and pacers or magnetic impurities in the lungs or in the head, etc. These problems, typical to patients, may severely reduce the clinical applicability of magnetoencephalography (MEG). The Signal Space Separation (SSS) method introduces a revolutionary improvement in feasibility and robustness of MEG measurements in the presence of such sources of artifacts. SSS allows recordings in environments with external interference of arbitrary geometry. With SSS, the compensation of subject-related artifacts typically arising from unintentional movements of the body and head is possible. With this method, every measurement can be uniquely decomposed into device-independent components with separate components for the interesting biomagnetic signals and external interference signals irrespective of the temporal or statistical nature of the sources. The device independence of this decomposition enables compensation of movement artifacts and analysis of DC phenomena.