摘要
提出一种针对心电信号特征检测的快速算法。首先针对干扰信号的频率特性分别采用了中值滤波、梳状滤波器、Butter worth滤波器去除噪声干扰。然后根据心电波形特性采用一种自适应阈值差分算法来进行QRS波检测。接着采用改进后的局域变换法检测P、T波。利用MIT-BIH的QT数据库对所提出的算法进行评估,QRS群波检测率达99.65%以上,对P、T波的定位误差在50 ms以内。试验结果表明,该算法能提高各波的检测精度,适于实际应用。
This paper describes a fast algorithm for feature detection of ECG. Firstly, the frequency characteristics of the interference signal with a median filter, comb filter, Butterworth filter is to remove interference. Then according to the ECG waveform characteristics adopt an adaptive threshold difference algorithm to detect QRS wave. Then use improved local transformation method to detect P,T wave. The data from the MIT-BIH QT database is validated to verify the improved algorithm, detection rate of the QRS complexes excelled to 99.65. The errors of the P-wave position and T-wave position are reduced under 50 milliseconds or less. Experimental results show that the improved algorithm improves the detection accuracy of the waves, accurate and suitable for practical application.
出处
《机电工程技术》
2013年第8期106-110,共5页
Mechanical & Electrical Engineering Technology
关键词
心电预处理
心电特征检测
自适应阈值差分法
局域变换法
pre-processing of ECG
ECG feature detection
adaptive slope threshold method
local transformation method
