Abstract
About three million people die from sudden cardiac arrest (SCA) or infarction each year. It is the biggest killer on the planet and it can happen to anyone, anywhere, and anytime. Early and effective diagnosis from ECG signals can help save a large part of these SCAs. Premature ventricular contraction (PVC) detection and classification was performed with different methods. A new method of automatic classification of PVC based on a three-bit linear interpolation error signal (LIES) is proposed in this paper. We start by a QRS detection, a nonlinear transformation and then a sliding window. This method was tested on normal and abnormal beats obtained from MIT-BIH and AHA databases. The proposed algorithm with a sensitivity 94.8% and a specificity was 98.6%. is accurate. This algorithm is also automatic, robust, and fast. (c) 2022 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.