Blind-Signal-Separation

  This paper presents a new algorithm to extract the fetal ECG from an ECG recorded on the mother’s abdomen. The algorithm consists of two steps: first, ECG of fetus is extracted from the original signal using a two-level wavelet transform. We use Daubechie wavelet transform, which is similar in shape to heart beat wave. The resulted signal is then low-pass filtered using the Savitzky-Golay smoothing filter to attenuate the effect of noise. The filtered signal is considered as the fetal ECG. The algorithm was applied on both synthetic and real ECG signals. The results in this paper show that the proposed algorithm has promising performance.