摘要
胎儿心电信号提取对胎儿监护具有重要意义。本文介绍了一种基于自适应线性神经网络的胎儿心电信号提取方法。该方法根据母体心电信号与母体腹部信号的相关性原理,以母体心电信号为网络输入,母体腹部信号为网络目标,采用W-H学习方法获取的训练误差即为提取出的胎儿心电信号。此外,通过增加网络隐含层,对神经网络的结构进行改进,增加网络训练精度,从而得到更好的训练结果,提取出更易识别的胎儿心电信号。最后分别使用仿真数据和临床数据对上述方法进行测试,实验结果表明,利用自适应线性神经网络可以提取出胎儿心电信号,通过改进神经网络结构,可以提取出更为清晰的胎儿心电信号。
Fetal ECG signals extraction has the vital significance for fetal monitoring. This paper introduces a method of extracting fetal ECG based on adaptive linear neural network. The method was due to the correlation between maternal ECG and the abdominal signals of pregnant woman, adopted the W-H learning rule, with maternal ECG as input signals and the abdominal ECG as target signals of network, so that the training error was obtained as the fetal ECG extracted. In addition, a better result was achieved by increasing the hidden layer for the network to improve neural network structure and enhance the network training accuracy. Thus, more easily identified fetal ECG would be extracted. Using simulated data and clinical data to test this method, experimental results showed that the adaptive linear neural network could be used to extract fetal ECG signals from maternal abdominal signals effectively. Furthermore, clearer fetal ECG signals could be extracted by improving neural network structure.
出处
《北京生物医学工程》
2010年第6期575-580,共6页
Beijing Biomedical Engineering
基金
北京市中青年骨干教师基金项目(102KB00845)资助
