Automated Detection of Posterior Myocardial Infarction From VCG Signals Using Stationary Wavelet Transform Based Features
Eedara Prabhakararao, Samarendra Dandapat
Abstract
Posterior myocardial infarction (PMI), also known as “the dark side of the moon,” is a lethal heart condition that can cause a heart attack if left untreated. The popularly used standard 12-lead electrocardiogram signals show poor sensitivity for the detection of PMI as it does not have posterior monitoring electrodes. The three-lead vectorcardiogram [(three-lead vectorcardiogram (VCG)] signals, on the other hand, has an electrode toward the posterior side, which improves its reliability for PMI diagnosis. Therefore, in this article, we exploit the three-lead VCG signals for the automatic identification of PMI patients from healthy control (HC) subjects. The proposed method quantifies the electrical conduction abnormalities of PMI patients by extracting discriminative multiscale eigenfeatures from the stationary wavelet transform subband matrices. Furthermore, to combat class imbalance, a cost-sensitive support vector machine classifier is used. The experimental results on the physikalisch-technische bundesanstalt (PTB) diagnostic database show an impressive PMI detection accuracy without compromising on the HC detection.