Analysis of high frequency fetal heart rate variability using empirical mode decomposition

Abstract

We used empirical mode decomposition (EMD) to assess the high frequency information of fetal HRV during different conditions of fetal activity. Beat-to-beat data of thirteen normal fetuses at term (36 to 40 gestation weeks) were obtained from abdominal ECG signals. In addition, corporal and breathing fetal movements were simultaneously observed using ultrasound B mode images. High frequency components of fetal HRV data were obtained by applying the EMD and by reconstructing the signals from components having frequencies higher than 0.3 Hz, according to the lowest frequency of fetal breathing movements observed in our data. Our results show significant differences in the power of high frequency components for episodes only with breathing movements vs no movements and vs. episodes having corporal movements (p<0.05). In conclusion, EMD seems to be a useful tool to analyze the effect of the FBM on non-stationary fetal HR time series