Independent component analysis of noninvasively recorded cortical magneticDC-fields in humans

We apply a recently del eloped multivariate statistical data analysis techn ique-so called blind source separation (BSS) by independent component analy sis-to process magnetoencephalogram recordings of near-de fields. The extra ction of near-de fields from MEG recordings has great relevance for medical applications since slowly varying de-phenomena have been found, e.g., in c erebral anoxia and spreading depression in animals, Comparing several BSS approaches, it turns out that an algorithm based on t emporal decorrelation successfully extracted a de-component which was induc ed in the auditory cortex by presentation of music. The task is challenging because of the limited amount of available data and the corruption by outl iers, which makes it an interesting real-world testbed for studying the rob ustness of ICA methods.