Analysis of Rest and Exercise-induced 3-D Magnetocardiogram and Body Surface Potential Map using Singular Value Decomposition

Abstract

Magnetocardiograms (MCG) represent a non-invasive method of measuring functional information concerning the electrical activities of the heart. We carried out exercise-induced MCGs and body surface potential maps (BSPMs) for normal subjects. The MCGs were measured with a three dimensional second-order gradiometer connected to a 39-channel superconducting quantum interface device (SQUID), which could simultaneously detect magnetic field components perpendicular to the chest wall (Bz) and tangential to the chest wall (Bx, By). The power map was calculated to examine the change in spatial distribution before and after exercise. We found that the spatial distribution at QRS complex was different before and after exercise when we applied singular value decomposition (SVD) to the power map. Time-frequency analysis (wavelet analysis) was applied to the MCGs and BSPMs to examine the change in the power spectrum of the QRS complex and ST segment. We found that the frequency component of Bz in MCGs reconstructed by the second singular value was different.