A New Adaptive Modeling and Denoising of Real ECG signal

Heart problems are increasing day by day and Electrocardiogram (ECG) signal is extremely important to evaluate cardiac problems. A common problem in ECG interpretation is noise removal because real ECG signals are corrupted by various types of noises and artifacts with different frequencies. Among all the noises the power line interference is the most significant and can strongly affect the ECG signal. This paper presents second order FIR Notch filter to remove power line interference which is occurred with frequency of 50Hz in India, followed by a novel parametric representation of real ECG signals. First, Notch filter is designed here for two cases first one is filter with two zeros, and in second case adding two poles at different radius but in same frequency as those of zeros, but the radius is less then unity in all the cases to consider the stability of filter. The effect of poles with their variables radius (0.8o to 0.99) is also analyzed at the output with SNR and has considered best result in term of stability, magnitude and phase response. Second part includes Signal acquisition, Control point detection, and develops a spline model of real ECG signal.