Measuring the Time-Scale-Dependent Information Flow Between Maternal and Fetal Heartbeats During the Third Trimester: Impact of Fetal Sex and Maternal Chronic Stress

Figures

• 

  Figure 4. Entropy rate h as a measure of information or complexity in the conditioned signals. (a) Results when conditioning is performed on the maternal HR signal, and (b) when it is performed on the fetal HR signal. Dotted lines pertain to fetal HR and continuous lines to maternal HR. Blue (resp., orange) curves are obtained by conditioning on mHR accelerations (resp., mHR decelerations). Green (resp., red) curves are obtained by conditioning on fHR accelerations (resp., fHR decelerations). For reference, we have plotted the entropy rate measured in fHR (dotted black line) and mHR (continuous black line) without any conditioning. For each curve, a shaded area represents the dispersion of the values over the cohort.

• 

  Figure 1. Recruitment flow chart for the FELICITy dataset. The flowchart depicts the original cohort of 164 screened participants (79 stressed, 85 controls). After removal of 47 dyads (45 due to fHR indistinguishable from mHR, 1 dyad lacking clinical records, and 1 dyad with a missing PSS score precluding stress classification), 118 mother–fetus dyads with complete clinical and entropy data were retained for analysis (59 stressed, 59 controls; 49 male, 69 female fetuses). One participant (FS-124) was additionally excluded from Table 4 only (n = 117) due to an extreme outlier net TE value (z-score = 23.7).

• 

  Figure 2. Statistics of deceleration/acceleration ratios computed at time scale τ=2.5 s. The upper and lower parts correspond to the mother’s HR and the fetus’s HR, respectively. In each case, the first line presents the fraction of time points that are decelerations or accelerations. In contrast, the second line presents the ratio of decelerations to accelerations.

• 

  Figure 3. Group dependencies of statistics of deceleration/acceleration ratios computed at time scale τ=2.5 s, for female (♀) and male (♂) fetuses, in both the stressed and control group.

• 

  Figure 5. Typical dependence of the maximal value hmax of entropy rate (left) as well as of its mean value hAUC (right) in the range [0.5–2.5] s for the mHR (blue—m) and the fHR (pink—f) across the cohort, as well as in each subgroup (stressed or control, female or male fetus).

• 

  Figure 6. Same as Figure 5 but when considering either accelerations of decelerations, computed on either maternal or fetal HR.

• 

  Figure 7. Typical dependence of the maximal value TEmax of the net transfer entropy from mother to fetus, as well as of its mean value TEAUC in the range [0.5–2.5] s across the cohort (all), as well as in each subgroup (stressed or control, female or male fetus). On the left subplots is also presented the components TEm→f (mother to fetus) and TEm→f (fetus to mother) to show their relative order of magnitude: the net TE is smaller than its one-way components.

• 

  Figure 8. Same as Figure 7 for net TE between the mother and the fetus, but when considering either accelerations of decelerations, computed on either maternal or fetal HR.