摘要
心电图波形特征提取是针对一维心电信号的弱信号特征提取。如何排除各种干扰,提取出心电波形特征,并准确定位心电信号中P波、QRS波群,T波,一直是心脏病智能诊断的难点和热点问题,其中QRS波群的定位又是其它波定位的重要依据,利用形态学和小波包理论相结合的方法对这一问题进行了探讨,提出了QRS波群定位和滤除基线漂移的方法。实验证明提出的方法速度较快,能较准确的定位QRS波群、有效的去除基线漂移。
The feature extraction ofECG waveform is a kind of weak signal characteristic extraction aiming at one dimension ECG signal. It is a difficulty and focused problem during diagnosing intellectively cardiopathy that how to exclude all kinds of interference and to extract the ECG waveform feature and to locate accurately P wave, QRS waves, T wave of ECG signals. Moreover, the location of QRS waves is the important foundation of other wave locations. This problem is discussed using a combined method based on morphology and wavelet packet theory and a method of wave location and filtering baseline excursion is posed. The experimental result prove to be faster and to locate accurately the QRS waves and to takeoff effectively baseline excursion.
出处
《计算机工程与设计》
CSCD
北大核心
2007年第11期2726-2729,共4页
Computer Engineering and Design
关键词
心电图
小波包
一维形态学
特征提取
基线漂移
ECG
wavelet packet
dimension morphological
feature extract
baseline excursion
