一种实时鲁棒的QRS波群检测算法

A REAL TIME AND ROBUST QRS COMPLEX DETECTION ALGORITHM

针对心电信号易受噪声影响,检测不准确的不足,提出一种鲁棒的QRS波群检测算法。算法首先采用小波变换去除基线漂移及高频噪声的影响,再利用差分方法检测心电信号中的局部极值点,根据R峰幅值高的特点利用幅度阈值检出信号中的候选R峰,最后根据信号小波域信息设定小波阈值删除多检R峰。经MIT-BIH心电数据库的验证,QRS波群检测的灵敏度为99.88%,正检测率为99.86%。实验结果表明,该方法简单高效、鲁棒性好,适于实时应用。

Since ECG is sensible to noise and is imprecise while detecting, in light of these, we propose a robust QRS complex detection algorithm. First, the impacts of the baseline shift and high frequency noise are removed using wavelet transform, then the local extreme points in ECG is detected with difference method, the candidate R peaks in the signal is obtained using amplitude threshold according to the characteristic of high amplitude of R peak, at last the wavelet threshold is set according to wavelet domain information to delete the redundant R peaks detection. Verified by the ECG database of MIT-BIH, the QRS complex detection reaches the sensitivity of 99.88% and the positive detection rate of 99.86%. Experimental result exhibits the simplicity, efficiency and robustness of the algorithm, it is suitable for real-time application.