Prenatal Stress Detection from Electrocardiography Using Self-Supervised Deep Learning: Development and External Validation

TL;DR

This work tackles the need for objective prenatal stress screening by applying self-supervised learning to pregnancy ECG signals. A ResNet-34 encoder pretrained with SimCLR learns multi-layer representations from maternal, fetal, and abdominal ECG, with a downstream logistic and ridge regression head for binary stress classification and continuous PSS-10 prediction. Multi-layer features substantially outperform single-layer and foundation-model embeddings, achieving near-perfect internal accuracy (e.g., mECG $\approx$99% and fECG $\approx$100% classification) and strong regression $R^2$ values ($0.88$–$0.96$), while fetal ECG provides the strongest signal. External validation on FELICITy 2 shows robust generalization despite hardware differences, with signal quality gating further boosting performance; importantly, the approach can detect intervention-related stress trajectory changes ($p=0.041$). These findings support the potential for continuous, objective prenatal stress monitoring and evaluation of stress-reduction strategies in clinical practice.

Abstract

Prenatal psychological stress affects 15-25% of pregnancies and increases risks of preterm birth, low birth weight, and adverse neurodevelopmental outcomes. Current screening relies on subjective questionnaires (PSS-10), limiting continuous monitoring. We developed deep learning models for stress detection from electrocardiography (ECG) using the FELICITy 1 cohort (151 pregnant women, 32-38 weeks gestation). A ResNet-34 encoder was pretrained via SimCLR contrastive learning on 40,692 ECG segments per subject. Multi-layer feature extraction enabled binary classification and continuous PSS prediction across maternal (mECG), fetal (fECG), and abdominal ECG (aECG). External validation used the FELICITy 2 RCT (28 subjects, different ECG device, yoga intervention vs. control). On FELICITy 1 (5-fold CV): mECG 98.6% accuracy (R2=0.88, MAE=1.90), fECG 99.8% (R2=0.95, MAE=1.19), aECG 95.5% (R2=0.75, MAE=2.80). External validation on FELICITy 2: mECG 77.3% accuracy (R2=0.62, MAE=3.54, AUC=0.826), aECG 63.6% (R2=0.29, AUC=0.705). Signal quality-based channel selection outperformed all-channel averaging (+12% R2 improvement). Mixed-effects models detected a significant intervention response (p=0.041). Self-supervised deep learning on pregnancy ECG enables accurate, objective stress assessment, with multi-layer feature extraction substantially outperforming single embedding approaches.

Figures (5)

• Figure 1: Model Architecture and Multi-Layer Feature Extraction Strategy.
• Figure 2: FELICITy 1 Classification and Regression Performance.
• Figure 3: FELICITy 2 External Validation Performance.
• Figure 4: Temporal Stress Trajectories and Intervention Effects.
• Figure 5: Signal Quality Impact on Prediction Accuracy.