QRS complex is the most important part in electrocardiogram (ECG) as it contains the most important information of heart activities. R-peak detection is the first, yet crucial, step in most ECG automatic diagnose methods. Due to the existence of noise in ECG signals and changes in QRS morphology, most existing methods are not robust in different conditions. In the field of intelligent remote health caring, in addition to the detection accuracy, timeliness is also an important research issue. In this paper, wavelet transform and energy window transform are introduced, which form the basis of a novel R-peak detection method. Wavelet transform is used to efficiently reduce noise and highlight useful ECG signal for it has good time-frequency resolution characters, and energy window transform converts time domain signal to energy domain, which makes it easier to isolate QRS complex from other signals. As a result, influence from QRS morphology changes can be effectively alleviated. To validate the effectiveness of this new method, ECG records of MIT-BIH arrhythmia database are used in the experiments. The experiment results show that the proposed method is efficient and robust to noise and QRS morphology changes. The computational cost of the proposed method has also been evaluated.
An Adaptive Peak Detection Method for Inspection of Breakages in Long Rails by Using Barker Coded UGW
