Electrocardiogram signal feature detection algorithm based on lifting wavelet and improved envelope
A method for electrocardiogram (ECG) signals denoising based on lifting scheme wavelet transform, Hilbert transform and an improved approximated envelope was proposed. In addition, a detection algorithm based on the slop of threshold was introduced. In this work, the wavelet coefficients were obtained by decomposition of lifting scheme wavelet transform. An improved half-soft threshold was used for ECG signal denoising. After the threshold denoising, Hilbert transform and improved approximated envelope were operated with the ECG signals. Then a detection algorithm based on the slop of threshold was used for feature detection. The results of simulation demonstrate that the method proposed enables us to detect the location of R peak, P peak, T peak, the start and end of QRS and premature ventricular contraction. Algorithm performance was evaluated against the MIT-BIH Arrhythmia Database and the numerical results indicate that it can function reliably even under the condition of poor signal quality and long P and T peaks. The R peak detection error rate of the tape 105 is merely 0.27%.