ECG signal analysis using autoregressive modelling with and without baseline wander

According to the report of Times of India, India is becoming the heart disease capital of the world. Consequently, its need of time to make more efforts to enhance the research regarding heart and heart disease to keep the country away from facing any unpleasant and disastrous situation in future. For correctly diagnosing the type of heart disease, Electrocardiogram (ECG) is an important tool which conveys the information in terms of three waves namely; P-wave, QRS-wave, and T-wave. For analyzing the patterns (characteristics) of these waves, Autoregressive (AR) coefficients are required. In this paper,various real recordings are done. For classification purpose, two techniques viz. K-Nearest Neighbor (KNN) and Principal Component analysis (PCA) are used individually. Autoregressive modelling is done on ECG signal with baseline wander (BLW) and ECG signal without BLW for comparing the performance.