Reconstruction of Premature Atrial Contraction and Premature Ventricular Contraction on ECG Traces by Applying PLA as Segmentation Process

Abstract

Premature Atrial Contraction (PAC) and Premature Ventricular Contraction (PVC) are two types of ECG arrhythmias that are identified for presenting anomalies in the normal cardiac rhythm, generating alterations in the heart rate that disrupts his mechanic and electric heart activity due to these delayed contractions (premature). Also when a delayed contraction occurs (i.e. PVC or PAC), it generates alterations in heart rate that can produce ventricular tachycardia and eventually sudden cardiac arrest. Piece Wise Linear Algorithm (PLA) is a widely used segmentation process that represents each signal as linear segments with different scopes. In this work we are presenting an evaluation of PLA and the optimal number of segments when this type of segmentation is applied on normal ECG traces and when these ECG traces present PAC or PVC. We established that optimal number of segments is related with the coefficient of determination (R 2), before and after the reconstruction of an ECG trace. Results show that number of segments taken on normal ECG traces differs radically compared when this method was applied on traces with PVC and PAC events.