Detection of ECG Fiducial Points using Moving Average Filters for Diagnosis of Cardiac Disease

This paper is an approach to develop an easy, effective and robust algorithm for the detection of ECG fiducial points and thereby using the information provided by these fiducial points for the analysis of the patient's cardiac health. This algorithm makes use of Adaptive Thresholding and Moving Average Filters for QRS Complex detection. The identification of P-wave and T-wave is carried out by a simple search forward and backward procedure. The algorithm has been tested on several ECG data available on online database, i.e. PhysioNet with sampling frequency ranging from 128Hz-1 KHz. The accuracy of this technique has been determined on the results obtained by two statistical parameters, i.e. Positive Predictivity (+P) and sensitivity (SE).