Independent Component Analysis for Fetal Heart Rate Detection Using Photoplethysmography

Photoplethysmography (PPG) is an optoelectronic technique for measuring and recording changes in the volume of body parts. These changes are associated with each heart beat and acquired by pulse oximetry. Fetal heart rate (FHR) monitoring using PPG is a challenging task since the acquired signals present the pattern of both fetal and maternal hearts. The effect of maternal component, noise and artifacts on fetal component makes the separation of FHR very difficult. In this paper, we study the applicability of independent component analysis (lCA) to FHR detection using PPG. The study was conducted using emulated signals to mimic the pulsation nature of both maternal and fetal hearts. The outcome of this experiment shows encouraging results in terms of the extraction ability oflCA, which can perform well even when fetal-to-maternal signal-to-noise ratio (SIY~) drops to -276 dB.