Discrimination of Multiple Sources of 3D Biomagnetic Field using Frequency and Phase Analyses

Abstract

This paper describes the usefulness of 3D biomagnetic field measurements for discriminating multiple sources closely located and overlapped in time such as in magnetocardiogram(MCG) and magnetoencephalogram (MEG), using frequency and phase analysis. 3D MCG measurements with 48 positions on the chest wall were carried out on normal subjects and on a patient with Wolff-Parkinson-White (WPW) syndrome using a 3D vector second-order gradiometer connected to 39-channel SQUIDs, which can simultaneously detect magnetic fields perpendicular (Bz) and tangential (Bx, By) to the chest wall. To discriminate multiple sources activating in the heart we developed a signal processing method consisting of singular value decomposition(SVD) to estimate the number of equivalent sources, time-frequency analysis to calculate the frequency spectrum of 3D MCG wave-forms, and coherency analysis to calculate phase coherency at each measurement position. We were able to discriminate multiple sources consisting of the source of an accessory pathway and the source of a normal pathway hidden by an intense signal such as the delta wave of the WPW patient, using coherency analysis with the tangential component (Bx, By) added to the normal component(Bz).