Preterm birth risk stratification through longitudinal heart rate and HRV monitoring in daily life

Abstract

Preterm birth (PTB) remains a global health concern, impacting neonatal mortality and lifelong health consequences. Traditional methods for estimating PTB rely on electronic health records or biomedical signals, limited to short-term assessments in clinical settings. Recent studies have leveraged wearable technologies for in-home maternal health monitoring, offering continuous assessment of maternal autonomic nervous system (ANS) activity and facilitating the exploration of PTB risk. In this paper, we conduct a longitudinal study to assess the risk of PTB by examining maternal ANS activity through heart rate (HR) and heart rate variability (HRV). To achieve this, we collect long-term raw photoplethysmogram (PPG) signals from 58 pregnant women (including seven preterm cases) from gestational weeks 12-15 to three months post-delivery using smartwatches in daily life settings. We employ a PPG processing pipeline to accurately extract HR and HRV, and an autoencoder machine learning model with SHAP analysis to generate explainable abnormality scores indicative of PTB risk. Our results reveal distinctive patterns in PTB abnormality scores during the second pregnancy trimester, indicating the potential for early PTB risk estimation. Moreover, we find that HR, average of interbeat intervals (AVNN), SD1SD2 ratio, and standard deviation of interbeat intervals (SDNN) emerge as significant PTB indicators.

Keywords: Health monitoring; Heart rate; Heart rate variability; PPG; Preterm birth.

Fig. 1

Data analysis workflow.

Fig. 2

Autoencoder architecture.

Fig. 3

Weekly average abnormality scores for 7 pregnant women with PTB and 7 pregnant women with term delivery during the second and third trimesters.

Fig. 4

Average of weekly average PTB abnormality scores of preterm and term users.

Fig. 5

Distribution of PTB abnormality score for term and preterm classes.

Fig. 6

PTB abnormality score of HR and HRV features for term and preterm samples.

Fig. 7

The importance of each HR and HRV feature in PTB abnormality score.

Fig. 8

Mean SHAP values of each HR and HRV feature for preterm samples.

Fig. 9

Dependence plot of four HR and HRV features with the highest SHAP value.

Fig. 10

Receiver operating characteristic curves of 5-fold cross-validation.