Comparison of Adaptive filters in extraction of Fetal ECG

Fetal ECG extraction during pregnancy is important to monitor the functioning of the fetal heart. Various abnormalities related to the heart can be detected by analyzing the ECG. Adaptive filters are being used to extract fetal ECG from acquired ECG signal which consists of both maternal and fetal ECG. This paper compares three different adaptive filters, LMS - least mean square, NLMS - normalized least mean square, and Kalman filter in its ability to extract fetal ECG from mixed signal. The comparison between characteristics such a signal to noise ratio, computational complexity, efficiency, and stability lead to a necessary inference. Fetal extraction methods were implemented on MATLAB and fetal ECG was successfully extracted.