一种改进的心电图QRS波群检测算法

在利用小波变换检测QRS波群时,最关键的部分就是模极值配对,提出一种区域极值配对算法来检测R波。首先利用二次样条小波基函数和多孔(ATrous)算法对心电(ECG)信号进行小波变换求取模极值,用正极大值来确定搜索区域,以这个正极大值为起点,以这个确定区域为搜索范围,向左搜索负极大值点,将这两个极值配对,他们之间的过零点就是R波的对应点,然后在检测到R波的基础上检测出Q波与S波,再结合距离最大值法检测出QRS波群的起止点。并采用医学相关理论对检测结果进行优化,进一步去除错检点,补偿漏检点。最后利用MIT-BIH心率失常数据库中记录的数据对该算法进行验证,实验结果表明所提算法能准确检测QRS波群,平均检出率达到了99.97%。

An improved electrocardiogram QRS complexes detection algorithm

While the wavelet transform is used to detect QRS complexes, the most critical part is the match of mould extreme value. We present a regional exetrem value matching algorithm to detect R waves. The wavelet transform is realized by means of the quadratic spline wavelet basis function and the Atrous algorithm, and then the mould extreme values of ECG signals are calculated. The search area is defined by the positive maximum. Taking the positive maximum as the starting point, the specific parts as the search space, we search for the negative maximum from the starting point to the left. The zero crossing point between the positive maximum and the negative maximum is the corresponding point of R wave. Then we detect the Q wave and S wave on the basis of the R wave detection, and the onset and offset points of QRS complexes are detected by the maximum distance method. Related medical theories are also adopted to optimize the detection results, thus removing wrong points and compensating leak points. Experiments on the MIT BIH arrhythmia database verify the proposed algorithm, which can detect the QRS complexes accurately with an average detection rate of 99.97%.