Wavelet decomposition of the signal averaged electrocardiogram used for the risk stratification after acute myocardial infarction

Wavelet decomposition (WD) is a time-scale technique suitable for the detection of small signals which are hidden in large waves. Here, the authors investigated the use of a WD of signal averaged electrocardiogram (SAECG) for stratification of 106 patients who suffered acute myocardial infarction. The WD of SAECG was employed to identify 53 patients who developed sustained ventricular tachycardia during follow-up and to distinguish them from 53 age, sex, and infarct site matched patients with a uncomplicated follow-up. The study showed that (a) the reproducibility of WD indices of SAECG is as high as that of the conventional time-domain indices, and (b) the WD based identification of patients with ventricular tachycardia, assessed by computing the receiver operator characteristics is similar to that based on the convectional time-domain indices. The stratification of a subpopulation with anterior infarction was more precise using the WD than using the conventional time-domain analysis. The study concludes that wavelet decomposition is an alternative way for the evaluation of SAECG recording.<>