Automatic detection of QRS onset in ECG signals

In this paper, a new modified method based on the multi-scale morphological derivative (MMD) transform technique is described for the detection of QRS onset in ECG signals. According to our experiments, we find the criterions of wave classification, the appropriate value of threshold and transform scales. When these are all settled down, this algorithm can set different scales to perform the MMD transform according to different lengths of QRS waves. By testing our method with ECG data from CSE database, we prove that our method can process lots of signals with various waveforms efficiently.