A Low Complexity Algorithm for Detection of Three types of Cardiac Arrhythmia

This paper explores a simple yet novel method for detection of three kinds of arrhythmias from Electrocardiography (ECG) signal taken from Lead II. There are mainly three stages to the process: ECG signal processing, ECG features extraction and actual detection of the arrhythmia. The processing of ECG unfolds in three levels of filtering followed by averaging. Important ECG features such as RR interval, heart rate, regularity of P and R waves, QRS duration and PR interval were extracted from the processed ECG signal. By a combination of these features and thresholding, the three arrhythmias: Premature ventricular contraction, Sinus bradycardia and Sinus tachycardia were detected. The novelty of the paper lies in the reduced computational overheads.