基于小波分析和神经网络的胎儿心电提取  

Extraction of FECG Based on Neural Network and Wavelet Analysis

作　　者：余尤好[1] 李文芳[1] 刘信禹[2] 

机构地区：[1]莆田学院电子信息工程系,福建莆田351100 [2]莆田学院附属医院检验科,福建莆田351100

出　　处：《莆田学院学报》2012年第5期83-87,共5页Journal of putian University

基　　金：福建省教育科学"十二五"规划2012年度常规课题(FJCGGJ12-034)

摘　　要：通过小波分析方法和自适应线性神经网络相结合,对围产期母体腹壁混合心电信号进行处理,采用两种方案进行仿真并分析对比。分别采用小波变换和小波包分解技术对心电信号消噪处理,探索一种提取出胎儿清晰心电信号的方法,为下一步胎儿心电信号特征提取和健康状况的诊断奠定基础。实验结果表明,先提取胎儿心电信号,再进行消噪处理效果较好。The wavelet analysis and adaptive linear neural network was combined to deal with and analyze the mixed ECG signa｜ on the abdominal of perinatal maternal. Two kinds of schemes were used for simulation and the results were analyzed and compared. Wavelet transform and wavelet packet decomposition technology were used in ECG signal denoising for clearly fetal ECG signal extraction. It laid foundation for the fetal ECG feature extraction and health status diagnosis. The experiment results showed that it was more effective to extract the feature of fetal ECG before denoising.

关 键 词：小波分析 神经网络 胎儿心电信号 自适应 

分 类 号：TN911.72[电子电信—通信与信息系统] R318[电子电信—信息与通信工程]