Modeling and Classification of the ST Segment Morphology for Enhanced Detection of Acute Myocardial Infarction

Abstract

A number of cardiac conditions such as acute pericarditis (PC) and early repolarization (ER) cause ST elevation which mimics ST-segment Elevation Myocardial Infarction (STEMI). Current guidelines recommend analyzing ST segment morphology to distinguish STEMI from these confounders. ST elevation in PC and ER (and possibly in STEMI) is concave (upward) in the JTpeak interval, while a convex or straight ECG ST segment is associated with the diagnosis of STEMI. We developed an algorithm to classify concavity characteristic of the ST segment. A quadratic polynomial regression algorithm was introduced to model the shape of JTpeak interval. Our diagnostic algorithm generated representative beats and measured the fiducial points and standard measurements such as ST level in 12-lead 10-sec segments of ECG recordings. JTpeak interval was modeled by a parabola using a least-squares polynomial regression algorithm. Classifier features such as curvature, parabola direction and vertex, model fit error, and the noise measure were determined. A bootstrap-aggregated tree ensemble classifier determined the ST segment shape. Our algorithm was evaluated on a 12-lead ECG database collected in two medical centers. Our ST segment polynomial regression model exhibited significant improvement in concavity detection versus a simple conventional method.