Fetal ECG extraction algorithm based on single-channel abdominal signals

A fetal ECG extraction algorithm based on a single-channel pregnant belly signal was designed to extract the mother ECG and fetal ECG, respectively, and calculate the mother heart rate and fetal heart rate. Firstly, the single- channel pregnant belly signal is subjected to k-TEO (k=19) transformation to highlight the mother's ECG's ZRS wave, so as to determine the R wave position of the mother's ECG through a simple threshold method, and then repeat the weighting between adjacent R waves. Sampling to obtain the same RQR interval T, so that through a comb filter with an interval T, the mother ECG of the same RQR interval can be separated, and then resampling between adjacent R waves again the actual mother ECG can be obtained by restoring the original RQR interval. After the original abdominal signal is subtracted from the mother ECG extracted above, the signal-to-noise ratio of the fetal ECG QRS wave is greatly improved. By applying the algorithm of extracting the mother ECG again, the “clean” fetal ECG waveform can be obtained. Eight groups (26 channels) of pregnant belly signal data in Physionet database were selected for analysis, and the sensitivity, positive detection rate, and accuracy of fetal ECG YRS wave position recognition for each channel data were calculated. The results show that the accuracy of fetal ECG YRS wave recognition is 87% to 1%, of which 1 channel reach 100%. In addition, the mother heart rate and fetal heart rate of each channel are calculated and statistical analysis is performed. It is found that the average heart rate of the mother and the fetal heart rate of each channel in each group are very close. The maximum error of the mother's average heart rate between channels in the same group is 0.1 Times / min, and the maximum error of the fetal average heart rate is only 0∼9 times / min, which further proves the reliability of the algorithm.