Fetal nerve maturation is a dynamic process, which is reflected in fetal movement and fetal heart rate (FHR) patterns. Classical FHR variability (fHRV) indices cannot fully reflect their complex interrelationship. This study aims to provide an alternative insight for fetal neural development by using the coupling analysis of uterine electromyography (UEMG) and FHR acceleration. We investigated 39 normal pregnancies with appropriate for gestational age (AGA) and 19 high-risk pregnancies with small for gestational age (SGA) at 28-39 weeks. The UEMG and FHR were recorded simultaneously by a trans-abdominal device during the night (10 p.m.-8 a.m.). Cross-wavelet analysis was used to characterize the dynamic relationship between FHR and UEMG. Subsequently, a UEMG-FHR coupling index (UFCI) was extracted from the multiscale coupling power spectrum. We examined the gestational-age dependency of UFCI by linear/quadratic regression models, and the ability to screen for SGA using binary logistic regression. Also, the performances of classical fHRV indices, including short-term variation (STV), averaged acceleration capacity (AAC), and averaged deceleration capacity (ADC), time- and frequency- domain indices, and multiscale entropy (MSE), were compared as references on the same recordings. The results showed that UFCI provided a stronger age predicting value with R2 = 0.480, in contrast to the best value among other fHRV indices with R2 = 0.335, by univariate regression models. Also, UFCI achieved superior performance for predicting SGA with the area under the curve (AUC) of 0.88, compared with 0.79 for best performance of other fHRV indices. The present results indicate that UFCI provides new information for early detection and comprehensive interpretation of intrauterine growth restriction in prenatal diagnosis, and helps improve the screening of SGA.
Keywords: cross-wavelet analysis; fetal heart rate; fetal neural development; intrauterine growth restriction; small for gestational age; uterine electromyography.