Fetal Electrocardiogram Analysis And Extraction Based On Nonlinear Time Series

The Fetal Electrocardiogram reflects the electrical activity of the fetal heart.It contains information on the health status of the fetus and therefore,an early diagnosis of any cardiac defects increases the effectiveness for the appropriate treatment. Usually,there are several technical problems associated with the extraction of FECG from signals recorded at the mother's abdominal surface.Nowadays,many approaches have been reported to get better FECGThis paper applies nonlinear time series analysis to the FECG analysis and extraction.Firstly,the theoretical foundation of nonlinear time series methods, including phase-space reconstruction,nonlinear geometrical invariant quantities was introduced.Then,the process of mixed ECG signal analysis was proposed. Furthermore,the system tests of nonlinearity and deterministic were used by surrogate data,recurrence plot and recurrence quantification analysis methods.Meanwhile,multiple sets of experiments from different aspects was taken to determine the proper embedding dimension and delay times,and then the phase-space reconstruction graph of mixed ECG signal was plotted.Moreover,the correlation dimension,the Kolmogorov entropy and the largest Lyapunov exponents were computed and the chaotic characteristic of mixed ECG signal was proved.Based on the property of single-lead recordings from abdomen of a pregnant woman,a locally linear projection technique to extract the fetal component was applied in this paper.The general planning and illustrated several important parameters in the algorithm were proposed.By performing the experiments and comparing with LMS,RLS and FastlCA methods,it shows that the local projection method is effective to separate the FECG from the maternal trace and other contaminations.