A Robust Approach for Detecting QRS Complexes of Electrocardiogram Signal with Different Morphologies

摘要

In this paper a robust approach for detecting QRS complexes and computing related R-R intervals of ECG signals named (RDQR) has been proposed. Tt reliably recognizes QRS complexes based on the deflection occurred between R & S waves as a large positive and negative amplitude differences in comparison with respect to other ECG signal (P and T) waves. The proposed detection approach applies the new direct algorithm applied on the entire ECG itself without any additional transform like (wavelet, cosine, Walsh transform, etc.). According to the strategy based on positive and negative deflection it overcomes the problem of QRS direction positive (upright) or negative (inverted). Three different types of ECG online database with duration of 10 sec (MIT-BTH Arrhythmia, ST Change Database and Normal Sinus Rhythm) are used to validate the detection performance. The results are demonstrated that the proposed detection approach achieved (100%) accuracy for QRS detection also very high accuracy in evaluating related R-R intervals.