Heart beat peak detection using signal filtering in ECG data

Current trends in health industry reveal that the major cause for death ratio increase is due to heart disease. Malfunctioning of heart leads to heart disease and there are multiple forms of heart disease. Electrocardiogram (ECG) is painless and basic test which can detect basic heart related problems. Heart beat variations can be identified by detecting heart beat peaks. Heart beat peak detection plays a vital role for efficient analysis of ECG signals. This paper deals with detecting heart beat peaks in noisy ECG signals which will be helpful to extract required features, to detect heart disease in earlier stage. ECG Signals are taken as primary input to detect the heartbeat peaks for feature extraction purpose. As the noisy EGC signal is normal due to distortion of the original ECG signal because of the various levels of noises, filtering the noisy ECG signal is necessary to detect the heartbeat peaks. Existing digital IIR filters such as Butterworth, Chebyshev Type I, Chebyshev Type II and Elliptic are commonly used for denoising ECG signals to retrieve sharp ECG signal waves. This paper is an attempt to apply the existing methodologies to Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) noise, stress test database ECG signals and perform a quantitative study based on the performance metrics such as specificity, sensitivity, accuracy, mean square error and signal to noise ratio. Both Chebyshev Type II and Elliptic filers reflected poor performance over this dataset. Hence this paper proposes a novel hybrid methodology called ButterChev which is a combined version of Butterworth and Chebyshev Type I filters. The proposed methodology resulted with improved performance metrics and it paves the way for better noise removal and peak detection for the given noisy ECG signal. The implementation process has been carried out using Matlab software environment.