Linear and Non-Linear Analysis of Heart Rate Variability in Premature Infants

Heart rate variability signal (HRV) was extracted from ECG recordings in a group of 26 premature babies (average gestational age 28.7 weeks). Data were collected at Columbia University Medical Center, NY. The aim was to identify possible risk conditions arising after premature birth. In order to analyze these signals methods based on Linear and Non-linear analysis were applied. Parameters from Time Domain, Frequency Domain, complexity Entropy measures and Phase Rectified Signal Averaging (PRSA) technique presented a significant separation between Active and Quiet sleep. Moreover, we applied parameters from fetal HR analysis, since prematurity is a transitory phase between fetal and mature baby condition. This novel multi-parametric approach will be the basis toward a classification tool based on autonomic system parameters during sleep to identify possible risks connected with prematurity.