提出一种新算法进行QT间期的自动分析方法对ECG信号进行小波变换,根据模极大值对的形态进行T波识别。结果,经MIT-BIH数据库检验,QRS波识别率达到了99.52%。经QT数据库检验,T波正确识别率达到了87.69%,与心电专家手工标注相比,误差为0.33ms±10.59ms。
Objective to present a new algorithm for automatic analysis of QT interval Methods ECG signal was first decomposed using the wavelet transform , then T wave recognition was done according to the module maxima pair's morphology. Results validated with MIT-BIH database, QRS detection rate was 99.52%. Validated with the QT database , the T wave detection rate was 87. 69%, and the difference between manual annotations and the detector's T wave ends was 0.33 ms±10.59 ms.