Acquisition, Feature Extraction and Classification of ECG Signals

In recent years, cardiovascular diseases are still one of the harmful diseases that threats to human life. Electrocardiogram (ECG) is a non-invasive technique used as a major diagnostic tool for monitoring heartbeats and detecting cardiovascular diseases. Therefore acquisition and analysis of ECG signal accurately has important meaning .Recently, there has been increased interest and demand in ECG measurement devices in the medical and research fields. Any kind of abnormality in cardiac beat is termed as arrhythmia which results variation in the shape of ECG signal. Detection of arrhythmia is mainly used in ECG abnormality detection for identifying heart related problems. But it’s difficult to find out the small variation in the ECG signal. So, a computer based diagnosis is required for the early detection of arrhythmias which can not distinguished precisely by human eye.ECG signal processing is the best method to overcome this problem. There are mainly two techniques in ECG signal processing. They are pre processing and feature extraction.ECG signals are obtained by after placing skin surface electrodes on the surface of the skin, it leads contamination of noises with ECG signals. Pre processing technique is used to filter out the ECG signal which was contaminated with noises. In this paper, Pan Tompkin’s method used for pre processing of ECG signal which eliminates the noises due to baseline wander and power line interferences. Wavelet transform approach is used to extract the features of each filtered ECG segment. Arrhythmia detection done after extracting the features from ECG signal .In this research ECG data were obtained from Physio bank ATM& raw ECG signal acquired from heart rate monitor later processed through harduino microcontroller. In this project Texas Instruments TMS320C6713 DSK is used to generate and process the electrocardiogram signal in digital domain