An ECG detection algorithm using wavelet and autocorrelation transform

This paper presents an accurate and efficient electrocardiogram (ECG) detection algorithm for wearable ECG devices used in body sensor network. Considering the quasi-periodic characteristics of the ECG signal, the proposed algorithm apply the wavelet and autocorrelation transform to simplify the self-adaptive threshold setting in the peak detection algorithm and exclude the unstable R-R wave bands which affect the accuracy of the heart-rate calculation to obtain the more accurate values. The experimental results show an average error-ratio of 0.356%. Comparing with results obtained in other literatures, this algorithm has a high practical value.