Abstract:

After analyzing the theory of wavelet transform's detection of signal singularity, we have modified a QRS wave detection method based on wavelet transform, which focuses on the interval mapping of R peak through observing the coefficients of wavelet decomposition, the omission of R peak, the retraction of R peak, the judgment of the forward R peak and the reverse R peak, and put forward a new kind of follow threshold. Using this algorithm,the QRS wave detection was experimented on 14 ECG(Electrocardiogram) data in MIT\|BIH database within the Matlab simulation environment. The experimental data shows that, with the use of the algorithm, the detection rate of the normal ECG and the abnormal ECG reaches 99.98%, which has high practical value.

Key words: QRS wave detection, electrocardiogram, wavelet transform, signal singularity