An algorithm based on Biorthogonal Spline Wavelet is developed for detecting ECG characters points. In the algorithm, the Biorthogonal Spline Wavelet Transform of the ECG signal is first calculated using Mallat Algorithm. And then the R-peak is located by finding the best modulus maximum pair of the wavelet transform in the scale of 23. A more robust method to find the modulus maximum pair is put forward in this paper. The methods finding the onsets and offsets of QRS complexes, P and T waves are also provided. The detection rate of QRS complexes for the algorithm is above 99.7% for MIT/BIH database and the processing time is considerably little.
展开▼