Characterization of the fetal diaphragmatic magnetomyogram and the effect of breathing movements on cardiac metrics of rate and variability

Abstract

Breathing movements are one of the earliest fetal motor behaviors to emerge and are a hallmark of fetal well-being. Fetal respiratory sinus arrhythmia (RSA) has been documented but efforts to quantify the influence of breathing on heart rate (HR) and heart rate variability (HRV) are difficult due to the episodic nature of fetal breathing activity. We used a dedicated fetal biomagnetometer to acquire the magnetocardiogram (MCG) between 36 and 38 weeks gestational age (GA). We identified and characterized a waveform observed in the raw data and independent component decomposition that we attribute to fetal diaphragmatic movements during breathing episodes. RSA and increased high frequency power in a time-frequency analysis of the IBI time-series was observed during fetal breathing periods. Using the diaphragmatic magnetomyogram (dMMG) as a marker, we compared time and frequency domain metrics of heart rate and heart rate variability between breathing and non-breathing epochs. Fetal breathing activity resulted in significantly lower HR, increased high frequency power, greater sympathovagal balance, increased short-term HRV and greater parasympathetic input relative to non-breathing episodes confirming the specificity of fetal breathing movements on parasympathetic cardiac influence. No significant differences between breathing and non-breathing epochs were found in two metrics reflecting total HRV or very low, low and intermediate frequency bands. Using the fetal dMMG as a marker, biomagnetometry can help to elucidate the electrophysiologic mechanisms associated with diaphragmatic motor function and may be used to study the longitudinal development of human fetal cardiac autonomic control and breathing activity.

Figure 1. Biomagnetometry raw data, ICA decomposition and Magnetic Field Maps with Corresponding Power Spectrum

Ten seconds of channel data (A.), the ICA decomposition (B.) and topographical magnetic field maps with their respective power spectra (C.) are shown. In figure 1A., the raw data are shown with R-peaks marked by colored triangles The fetal MCG (blue) can be seen imbedded in the maternal MCG (green). Note the undulating baseline in channel MRK1. In section B., the individual ICA components are displayed. Maternal (green) QRS is seen in components 1 and 4. Fetal QRS (blue) is seen in components 2 and 3. The fetal breathing component (red) is in component 5. The breathing component was superimposed over the same time period in raw data (Panel A) to demonstrate that the undulating baseline seen in channel MRK1 is the waveform generated by fetal breathing movement. In section C., the magnetic field maps of the primary components of interest are shown. The magnetic field distribution for the maternal cardiac component is typical of the maternal heart positioned above the sensor array. The fetal cardiac component map shows a dipolar source. The power spectra for both cardiac components are complex with multiple harmonics. The map for the fetal diaphragm is dipolar and orthogonal to the fetal cardiac map. The power spectrum for this component has a single peak consistent with the temporal frequency (~1.2 Hz) in component 7.

Figure 2. Four examples of fetal dMMG

Each section represents 360 seconds of dMMG in 4 individual cases to demonstrate the various morphologies and patterns of fetal dMMG activity. The first section shows data from a 38 week fetus, the remaining 3 blocks are from 36 week fetuses.

Figure 3. Time-frequency analysis, fetal dMMG and IBI time-series

Time (X-axis, 0-360 seconds) by frequency (Y-axis, 0.02-1.7 Hz) is shown in the top panel. Power (dB) is represented on the scale to the right. The fetal dMMG is compressed in the middle panel for the 360second (6 min) period and decompressed in the lower panel so that the specific waveforms can be compared to the compressed sample. Colored asterisks mark periods noted in the text to allow comparison of the dMMG between the compressed and decompressed samples. During periods of fetal breathing (oscillations in the dMMG), power increases in the band between 0.4 and 1.7 Hz. Concurrent with the increased high-frequency power is the presence of fetal RSA seen in the IBI time-series (expanded in the inset).

Figure 4. Fetaland AdultMCG, dMMG and Concurrent IBI Time-series

The fetal MCG is shown in the top panel (left). The bottom panel shows the IBItime-series(green) between 45-90 seconds. Fetal dMMG(red)is plotted above the IBItime-series to show the presence of fetal RSA during the breathing epoch. At 65 seconds, fetal heart rate begins to accelerate and the breathing epoch ends at ~72 seconds. RSA is not seen in the IBI time-series after 72 seconds. For comparison, the right panel demonstrates a paced breathing exercise in an adult, non-pregnant subject. In this case, the dMMG (red) is shown below the IBI time-series (green) and the resulting RSA is seen in the IBI time-series.