Fetal electrocardiogram extraction using support vector regressions and empirical mode decomposition

A novel method based on support vector regressions (SVRs) was proposed to extract the fetal electrocardiogram (FECG) from the abdominal composite signal of the pregnant woman. The proposed method employed two leads to separately collect the maternal electrocardiogram (MECG) at the thoracic area and the abdominal composite signal at the abdominal area of the pregnant woman. The MECG component in the abdominal composite signal is a nonlinear transformation of the MECG and the nonlinear transformation was identified by SVRs with limited samples. An optimal estimation of the MECG component in the abdominal composite signal was obtained by the MECG undergoing the nonlinear transformation. Then the FECG can be extracted by subtracting the optimal estimate of the MECG component in the abdominal composite signal. Finally the satisfied FECG can be obtained by suppressing the baseline shift and other additive noise in the extracted FECG using empirical mode decomposition (EMD). Visual results obtained from the real electrocardiogram (ECG) signals demonstrate the validity of the proposed method even when the fetal QRS wave was entirely overlapped with the maternal QRS wave in the abdominal composite signal. Copyright © 2009 Binary Information Press.