A novel method of FECG extraction combined self-correlation analysis with ICA

Abstract

The extraction of fetal electrocardiogram (FECG) signal has an important value in clinical application. The FECG is extracted from the maternal abdominal signal which is collected by several electrodes placed on the maternal abdomen. The traditional independent component analysis (ICA) model does not consider the temporal correlation in the process of separating maternal abdomenal signal. In this paper, a new method for extracting FECG is proposed, which combines self-correlation analysis with independent component analysis. Firstly, the self-correlation analysis is used for intercepting signals, which can decrease the temporal correlation. Then FastICA is applied to obtain the model parameters of ICA model. Finally, bring the mixed-signal into this model to extract the fetal ECG. The experiments are conducted by clinical signals. The results indicate that the method proposed in this paper could extract FECG well. This method is superior to traditional FastICA.