Automatic MCG Signal Quality Evaluation Using the Magnetocardiography and Electrocardiograph

Abstract

The poor Magnetocardiogram (MCG) signal easily affects the reliability of some important parameters derived from MCG signal, such as RR interval, QT interval and ST interval et al, and is vulnerable to trigger false alarms, thus the overload of alarms can lead to alarm fatigue and eventually the care staffs' desensitization. So a real-time MCG signal quality evaluation algorithm is important and can enable the inexperienced care staffs to collect MCG signal of sufficient diagnostic quality. In this work, we proposed a automatic MCG signal quality evaluation algorithm based on the multi-parameters calculated from the MCG data and ECG data, which are collected at the same time. Unlike the existing methods for evaluating the signal quality, 7 parameters were extracted from the MCG and ECG data. Thereafter, each parameter was divided into two levels (good or bad) according to the set threshold. Finally, the quality of the MCG signal is evaluated according to the majority rule. The performance of the proposed algorithm was evaluated using the data collected by the 36-channel MCG system. The classification accuracy achieved 91%. The lower complexity and high classification accuracy make the proposed algorithm to be applied to real-time MCG signal quality evaluation system.