Filter based enhancement of QRS complex of electrocardiogram

A specialized filter very similar to comb filter is used in this study to suppress the parts other than QRS complex of electrocardiogram (ECG) thereby enhancing the QRS complex. Using appropriate threshold and some decision rules the location of the QRS complex can easily be found. The proposed technique is used to detect the QRS complexes of normal as well as premature ventricular contraction, paced and left bundle branch block beats. A total of 31 ECG data sets of MIT-BIH database, each of 30 minutes duration is subjected for detection. An average successful detection rate of 99.89% is found for the QRS complex. The proposed method is also very fast. To process ECG of 30-minute duration it takes only an average of 1.5 seconds when implemented on a Pentium II, 350 MHz computer in MATLAB 5.3 environment.