Identifying fetal breathing movements on the basis of spectral analysis in fetal heart period time series

摘要

The initiation of fetal breathing movements in the second trimester of pregnancy is a sign of maturation and development. Fetal magnetocardiography permits the determination of high resolution fetal heart period time series. We have developed a procedure which exploits the effect of respiratory sinus arrhythmia (RSA) on such time series and permits identification of periods of breathing movements. Magneto-cardiograms were recorded in fetuses in the second and third trimester of pregnancy. Signal processing using templates and correlation analysis were applied in order to identify fetal R peaks (1 ms precision) and the corresponding RR interval time series were constructed. A program was developed which reads in fetal RR time series and displays them in both the time and frequency domain. The plot of the RR interval time series over time permits visual identification of possible fetal RSA. The display of the frequency spectrum of the complete time series between 0-2 Hz indicates whether dominant, persistent breathing movements are present. As the episodes of breathing movements may be only a few seconds short, the frequency spectrum can be calculated for subseries within user-specified time windows. The spectra of the subseries can be viewed singly (2-D) and as a function of time (3-D). Fetal breathing movements may be presumed when the subseries spectra consistently demonstrate a peak at a specific frequency between 0.4 and 1.4 Hz over episodes with a duration > 5 s. We present a number of fetal RR interval time series acquired from different gestational ages in which episodes of fetal breathing movements can be identified persisting for various durations at different frequencies. Comparing these with data without such spectral characteristics indicate that this data analysis approach is suited for the identification of fetal breathing movements by examining the frequency spectra of the heart period time series.