Analysis of spatial variation in the atrial fibrillation frequency from the multi-channel magnetocardiogram

During a magnetocardiography (MCG) study, a patient was recorded while in atrial fibrillation (AF), and we used this opportunity to conduct a pilot study of the spatial characteristics of the AF signal. Forty nine MCG channels were recorded. Atrial activity was isolated by the selection of a QRST-free section of AF. Fourier analysis identified the frequency characteristics of the atrial signal. Principal component analysis (PCA) was used to identify independent components in the atrial signal. Several discrete frequency peaks in the range 4 to 9 Hz were observed which were spatially grouped. PCA separated these frequencies into discrete components. The characteristics of the atrial signals may be associated with different re-entrant pathways.