Tian Xuelong, Yan Chunhong, Yu Yaqing, Wang Tianxing
College of Bioengineering and Key Lab for Biomechanics & Tissue Engineering under the State Ministry of Education, Chongqing University, Chongqing 400044, China.
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2006 Apr;23(2):257-61.
The detection of R-wave of ECG is essential to the analysis of the heart rate variability (HRV). In this paper, an R-wave detection method using wavelet transform(WT) is presented in line with the principle of discrete wavelet transform(DWT) and multi-resolution technique (MRT). We made use of the special properties of dbl wavelet in time-domain, decomposed the original ECG signals into 3-level detailed signals on different frequency bands by using DWT with Mallat algorithm, and got appropriate threshold values in different high frequency bands to distinguish R-wave. It is concluded that the algorithm had significant effects on it, which is verified by MIT/BIH (Massachusetts Institute of Technology/Boston's Beth Israel Hospital) ECG Database. The results show that R-wave could be detected accurately and localized precisely by this method, even when the patient was seriously sick or the signal was disturbed by noise. Consequently the method has a quite high locating precision (its error is not more than two sampled points and about 85 percent of the points of R-wave in ECG signal are localized precisely) and the correct detection rate of R-wave is 99.8% by using wavelet transform, so this method is quite feasible.
心电图R波的检测对于心率变异性（HRV）分析至关重要。本文依据离散小波变换（DWT）原理和多分辨率技术（MRT），提出了一种基于小波变换（WT）的R波检测方法。我们利用db1小波在时域的特殊性质，采用Mallat算法通过DWT将原始心电信号分解为不同频段的3级细节信号，并在不同高频段获取合适的阈值来区分R波。结果表明该算法对此效果显著，经麻省理工学院/波士顿贝斯以色列女执事医疗中心（MIT/BIH）心电图数据库验证。结果显示，即使患者病情严重或信号受到噪声干扰，该方法也能准确检测出R波并精确进行定位。因此，该方法具有相当高的定位精度（误差不超过两个采样点，且心电信号中约85%的R波点能被精确定位），采用小波变换时R波的正确检测率为99.8%，所以该方法相当可行。
