Fetal electrocardiogram extraction based on non-stationary ICA and wavelet denoising

Abstract

Fetal electrocardiogram (fECG) monitoring is a technique for obtaining important information about the condition of the fetus during pregnancy and labour by measuring electrical signals generated by the fetal heart as measured from multi-channel potential recordings on the mother body surface. It is shown in this paper that the fetal ECG can be reconstructed by means of higher order statistical tools exploiting ECG non-stationarity associated with post-denoising with wavelets. The method is illustrated on real fetal ECG data.

ICA-Based Fetal Monitoring

“…Unlike in the previous experiment, the fECG is visible in the third plot from the top, and the fetal heart rate can be estimated even though the signal-to-noise ratio is low. Further denoising may necessary using other techniques -see, e.g., ; Vigneron et al (2003)-but this is beyond the scope of the present Chapter. …”

“…For example, they include the contrast maximization (CoM2) method (Comon, 1994), JADE (Cardoso & Souloumiac, 1993), INFOMAX (Bell & Sejnowski, 1995), FastICA (Hyvärinen, 1999), Barros' method (Barros & Cichocki, 2001;Li & Yi, 2008), SOBI (Belouchrani et al, 1997), Pearson-ICA (Karvanen et al, 2000) or MERMAID (Marossero et al, 2003). ICA has been also used in combination with Wavelet transforms (Azzerboni et al, 2005;Vigneron et al, 2003), singular value decompositions (Gao et al, 2003) and neural networks (Yu & Chou, 2008), to cite some few examples. For a review of non-ICA based approaches, see, e.g.…”

Real-Time Fetal Heart Monitoring in Biomagnetic Measurements Using Adaptive Real-Time ICA

“…The signal-to-noise ratio of the extracted components can be improved by wavelet denoising [10]. Wavelet denoising is based on wavelet transformation, which can be applied to transform a signal from time domain into wavelet domain [25], i.e., wavelet coefficients.…”

Fetal ECG Extraction by Extended State Kalman Filtering Based on Single-Channel Recordings

“…These methods are based on the assumption of independent components (or more generally independent subspaces [8] or partitions [9]) for the maternal and fetal signals, or of the existence of some temporal structure for the desired signals [10], [11], [12]. In [13], [14], wavelet decomposition was also combined with blind source separation for extracting and denoising fECG signals. In another recent work, a new technique was proposed to fasten traditional Independent Component Analysis (ICA) method [15].…”