A review of different ECG classification/detection techniques for improved medical applications

Electrocardiogram (ECG) is an important diagnostic tool in medical engineering, presented in the form of electrical signal. Its complete analysis requires three stages viz. pre-processing, feature extraction, and classification/detection. The last stage provides the final outcome; hence, its careful selection is very important. Unfortunately, due to diverse and widely varied data management practices, no single technique is absolutely preferred over others. Therefore, an extensive literature survey of various classification/detection techniques is presented, and their effects are summarized. Also, different techniques related to ECG arrhythmia classification and detection are proposed and evaluated on-the-basis-of figure-of-merits (FoMs) for improved medical applications. This proposed technique is important to extract important clinical/pathological attributes of the ECG signals. It ensures the novelty in the biomedical signal processing (BSP). In this article different existing techniques are compared in the light of proposed techniques for various ECG arrhythmias classification and detection.