Extraction Of Fetal Electrocardiogram Based On Fuzzy Neural Network

Fetal ECG is very important for clinical diagnosis of fetal development and the fetal heart disease during pregnancy. Using the indirect method to extract fetal ECG is noninvasive and operated easily. However, fetal ECG is very weak, and often subject to a variety of serious noise interference extracted from the mother surface. Though many scholars had a great deal of research, an effective noninvasive extraction is still a problem. To deal with the problems above, the paper established a model of fetal ECG to synthesize the dynamic fetal ECG as a signal for evaluation, and then studied the fetal ECG extraction algorithm based on fuzzy neural network. The focus content and results are as follows:â‘ To synthesize dynamic fetal ECG and maternal ECG. We analyze the fetal ECG parameters using direct extraction, in accordance with the dynamic ECG model proposed by McSharry et al, respectively synthesize the dynamic fetal ECG and maternal ECG, and then adjust the ECG parameters to a reasonable medical statistics extent, obtaining a quasi-original signal library similar real signal. Simulation results show that the parameters of the synthetic ECG are basically matched with the actual parameters. Which lays a foundation to synthesize the mother mixed abdominal signal and evaluate the extraction effect of different algorithms at different fetal to maternal SNR.â‘¡To propose the method based on fuzzy neural network extracting the fetal ECG. We divide the abdominal mixed signal into two parts that one is the fetal ECG containing the noise, the other is the maternal ECG component in the abdominal mixed signal, assuming that it is a transform of the maternal ECG for spreading from the heart to abdomen. Using fuzzy neural network to learn mapping between the mother ECG and abdominal mixed signal, in order to obtain the maternal ECG component in the maternal abdominal signal, and then subtract the maternal ECG component from the maternal abdominal signal we will obtain the fetal ECG with noise. Simulation results show that the algorithm can use the simple two-lead to extract the complete fetal ECG successfully.