Characterization of FECG Signals to Uncover the Complexity of Fetal Heart Rate

In this chapter, we study the fetal heart rate from abdominal signals using multifractal spectra and fractal analysis. There exist many studies that discovered for cardiac diseases the cardiac rhythm displays self affine monofractal properties and its complexity and the multifractal structures have been regulated by neuroautonomic control mechanisms. The Fetal electrocardiogram (FECG) signal may provide precisely detailed information that could help clinicians make more timely and suitable decisions during labour. The primary factor driving interest in FECG signal analysis is its potential for medical diagnostics and applications. Fetal monitoring is increasingly requiring the extraction and detection of the FECG signal from composite abdominal data using strong and sophisticated techniques. We use the Abdominal and Direct Fetal Electrocar-diogram Database contains multichannel FECG recordings obtained from 5 different women in labor, between 38 and 41 weeks of gestation. On these five FECG recordings, we use autocorrelation or power spectral densities (PSD) analysis to determine whether the signal of interest exhibits a power-law PSD and to estimate the exponent from realisations of these processes. We use multi-fractal analysis to see if distinct statistical moments at various scales of these FECG signals exhibit any kind of power-law scaling. We plot the multi-fractal spectra of this database to compare the width of the scaling exponent for each spectrum. A quantitative analysis commonly known as the Fractal Dimension (FD) using the Higuchi algorithm has been carried out to illustrate the fractal complexity of input signals. Our finding shows that the fractal geometry can be used as a mathematical model and computational framework to further analysis and classification of clinical database. This study indicates that fractal dimension can be used as a complexity index for FECG recordings but needs further analysis to find a threshold for clinical studies to be used as a biomarker and diagnosis tool in these types studies.