ECG feature extraction and classification of anteroseptal myocardial infarction and normal subjects using discrete wavelet transform

In this paper, a novel methodology, based on discrete wavelet transform (DWT) is developed for extraction of characteristic features from twelve - lead Electrocardiogram recordings. The first step of this method is to denoise the signal using DWT technique. A multiresolution approach along with thresholding is used for the detection of R - Peaks in each cardiac beats. Followed, by this other fiducial points (Q and S) are detected and QRS onset and offset points are identified. Baseline is also detected and heights of R, Q, S waves are calculated. This, algorithm was validated using PTB diagnostic database giving a sensitivity of 99.6% and MITDB Arrhythmia, giving a sensitivity of 99.8%. The QRS vectors are calculated for normal and patients with Anteroseptal MI and a comparative study is presented. Accordingly, it has been found that classification of normal and AS MI is possible by computing the QRS vector. And a simple classification rule is established for this purpose.