Characterizing Fetal Sympatho-Vagal Balance through Multivariate Time-Varying Autoregressive Modeling of Magnetocardiografic Data
We propose a method to analyze fetal beat-to-beat heart rate variability (HRV) obtained from magnetocardiographic (MCG) measurements at different times of gestation using a multivariate time-varying autoregressive (MTVAR) model. Our approach is based on treating a group of HRV signals from a single fetus and measured at different gestational periods as a dynamical system, from which a suitable MTVAR model is obtained. Then, changes in the values of the coefficients of the MTVAR model are associated to changes in fetal sympatho-vagal balance. Such association is made by comparing the dynamics of the MTVAR coefficients to the SDNN/RMSSD ratio, which is known to be a potential marker for sympatho-vagal balance. Furthermore, an overall behavior of the sympatho-vagal balance during the gestation is estimated from the average values of the MTVAR coefficients and compared to the tendencies previously reported in the literature. In order to demonstrate the applicability of the proposed modeling approach, real magnetocardiographic data from a healthy fetus is analyzed. Preliminary results show that the MTVAR model provides good insight into the sympatho-vagal dynamics with greater specificity compared to traditional measures.
KeywordsFetal magnetocardiography sympatho-vagal balance heart rate variability time-varying autoregressive modeling
Unable to display preview. Download preview PDF.