FECG Extraction Using 1D Convolution Neural Network

The fetal electrocardiogram (FECG) signal traces the electrophysiological behavior of fetal. The signal is utilized to find the condition of a fetus during pregnancy and labor. The amplitude of these signals is often contaminated by noises. Noises change the morphology of FECG signals which prevent correct diagnosis. Nowadays, deep learning algorithms are gaining popularity in the bio-signal analysis. In this paper, the concept of convolutional neural networks (CNN) of deep learning models is utilized to extract FECG signals from noisy environments. The proposed model is tested with MIT-BIH signals, and the results obtained are found to be useful.