Electrocardiogram (ECG) Signals Feature Extraction and Classification using Various Signal Analysis Techniques

ECG signal shows the electrical activity of the hearts. These signals are non-stationary; they display a fractal like self-similarity. It is one of the most important physiological parameter, which is being extensively used for knowing the state of cardiac patients. They may contain indicators of heart disease, or even warnings about impending diseases. The indicators may be present at all times or may occur at random. Soft-Computing approach is an important tool in which two or more successive ECG recordings are compared in order to find disorders in cardiac. This paper presents the method to analyze ECG signal extract features and classification according to different arrhythmias. Cardiac arrhythmias which are found are Normal Sinus, Supraventricular Tachycardia, Right Bundle Branch Block, Left Bundle Branch Block, Ventricular Tachycardia. A dataset was obtained from a records set which were manually classified using MIT-BIH Arrhythmia Database Directory then features are extracted using DWT (Discrete wavelet transform) and classification is done according using various methods ANN (Artificial neural network), ANFIS (adaptive neuro-fuzzy inference system), SVM (State vector machine), & Statistical classifier.