Detection and estimation of T wave alternans using multilead analysis of ECG signal

There is a direct correlation between electrical changes in depolarization phase of cardiac cycle and increased risk of ventricular arrhythmia as well as sudden cardiac death, so detection and evaluation of these changes, named as T wave alternans (TWA), can provide new facilities for the physicians. However, exact detection of TWA, because of its small amplitude (sometimes smaller than the noise level) and fusing with biological noises, such as electrodes motion, muscles activity and breathing, is difficult. In this paper for detection of T-wave alternans, unlike conventional methods, we used a multilead method. The proposed method at first, applies a principal component analysis method to pre-processed signals; then by applying correlation method to the modified data detects T wave alternans. This method, in addition to accurate detection of T wave alternans, unlike other existing methods, can detect the location of the alternans.