An algebraic derivative-based method for R wave detection

Abstract

In this paper, a new robust method for R wave detection in ECG signal is proposed by using algebraic derivative estimation based technique. In fact, this new and efficient method relies on differential algebra, non-commutative algebra together with operational calculus. This technique allows noisy signal to be filtered via iterated time integrals and R wave slope to be emphasized. The ECG signal is then Hilbert transformed to be enhanced and threshold compared. The performance of the algorithm was tested by using the annotated records of MIT-BIH Arrhythmia Database. The robustness of the proposed R wave detector in presence of noise was also tested according to records from MIT-BIH Noise Stress Test Database. The overall performance is quite good even SNR as low as 6 dB.