摘要
本文采用小波变换和频谱标测技术从信号平均心电图中提取一组含有晚电位信息的比值特征参数，利用人工神经网络ＢＰ算法，完成对心室晚电位的检测。研究结果表明，采用神经网络检测晚电位的方法不仅具有较高的检测正确率，而且可以放松对信号平均心电图分析时段选取及ＱＲＳ终点定位准确度的要求。
Improved wavelet transform and spectromapping were applied to extract a group of feature parameters from signals of averaged ECG. Ventricular late potentials(VLP) was detected using propagation(BP) neural networks. The results of research showed that the detection of VLP by the method of artificial neural networks offered promising result being not necessary to know the particular duration of VLP and the particular offset point of QRS.
出处
《北京生物医学工程》
1999年第3期154-158,共5页
Beijing Biomedical Engineering
关键词
小波变换
人工神经网络
信号平均心电图
VLP
Ventricular late potetials(VLP)
Artifical neural networks(ANN)
Wavelet transform(WT)
Signal average electrocardiography(SAECG)
