MCG Analysis with Independent Component Analysis Based on 3-D Measurements
Kenichi Yamazaki, Koichiro KOBAYASHI, M. Ogoshi, B. S. Kim, M. Yoshizawa, Kenta Nakai, Y. Uchikawa · Journal of the Magnetics Society of Japan · 2004
We carried out three-dimensional vector measurement of magnetocardiograms (MCGs) for normal subjects and a subject with Wolff-Parkinson-White (WPW) syndrome, using a three-dimensional second-order gradiometer connected to 39-channel SQUIDs, which can simultaneously detect magnetic field components perpendicular to the chest wall (Bz) and those tangential to the chest wall (Bx, By). We then analyzed the MCGs by using independent component analysis (ICA).ICA is a useful method for separating independent signals from overlapping signals. We succeeded in carrying out dynamic ICA by changing the analytical interval timewise, and in separating the excitement due to WPW syndrome patient’s accessory pathway and normal pathway.