Multistage Adaptive filter for ECG signal processing

In this busy and competitive world, though we have a lot of affection with our parents and loved ones, we can not always be with them. The advancement in the field of information technology explores new paths for healthcare and E-health systems are becoming increasingly popular. The American company Biotelemetry provides healthcare facilities wirelessly to approximately one million patients. It was the first to provide real time heartbeat ECG monitoring, analysis and response to the patients at home or anywhere. In automatic ECG analysis or E-health Systems, The analysis of the ECG wave is done automatically, instead of a doctor so a noisy ECG signal will lead to wrong clinical diagnosis. During data acquisition or transmission, several types of artifacts are embedded in the signal which degrades the signal quality and the signal does not remain faithful for clinical diagnosis. For automatic ECG analysis these artifacts should be suppressed. Many techniques are found in the literature to remove these artifacts like wavelet transform, Neural Networks and Adaptive filters. In this paper we propose novel adaptive algorithm for processing of ECG signal. A single stage adaptive filter can remove only one type of artifact from the ECG because it has only one reference input. The proposed multistage filter can eliminate multiple types of artifacts because of multiple reference inputs so it can be applied to the systems where multiple types of interference are found, provided that the apriori knowledge about the interference is available.