Abstract: This paper presents an algorithm for QRS detection using biorthogonal wavelet transform and Hilbert transform. First, this algorithm eliminates the high frequency noise and highlights the R peak position through biorthogonal wavelet transform decomposition and reconstruction, and constructs the detection layer which is beneficial to QRS waves detection. Then a technique with differentiation and Hilbert transform is applied on the re-composed signal in order to suppress the effects of P and T waves and decrease the low frequency noise. Finally, R peak is detected according to the adaptive threshold and decision rules. The MIT-BIH arrhythmia dataset is used to verify the performance of the detection method. The accuracy of QRS wave detection results is 99.01%, and the algorithm has good robustness and real-time performance.

Key words: QRS waves detection, ECG, wavelet transform, differentiation, Hilbert transform, adaptive threshold