Integration of angiogenic pathway-related proteomics and clinical features via ensemble learning for…

Case-controlEuropean journal of obstetrics & gynecology and reproductive biology: X· Sep 2026

Integration of angiogenic pathway-related proteomics and clinical features via ensemble learning for precision risk stratification of early-onset preeclampsia.

Huan Yang, Jing Ma, Hui Guo … Li Shi

Maternal-Fetal MedicineCase-controlDiagnosisQ2IF 2.2

Ensemble machine-learning model accurately predicts early-onset preeclampsia risk stratification

Integration of angiogenic pathway-related proteomics and clinical features via ensemble learning for precision risk stratification of early-onset preeclampsia.

Huan Yang … Li Shi

European journal of obstetrics & gynecology and reproductive biology: X · Sep 2026 · Q2 in Obstetrics & Gynecology

Background

Early-onset preeclampsia (early-onset preeclampsia) is a severe pregnancy complication associated with significant maternal and neonatal morbidity.

Purpose

This study aimed to develop a precision risk stratification framework for early-onset preeclampsia by integrating angiogenic pathway-related proteomics with clinical features using an ensemble learning approach.

Methods

A prospective case-control study was conducted involving 240 pregnant women (120 early-onset preeclampsia and 120 healthy controls).

n = 240 pregnant women

Results

correctly separates high from low risk almost every time, far better than the old ratio

AUC 0.95 (95% CI 0.92 to 0.98)

null 0

0.92

0.98

CI excludes the null - significant

More results

SHAP analysis identified the sFlt-1/PlGF ratio and Endoglin as the top contributors, revealing complex non-linear interactions between circulating biomarkers and maternal hemodynamics.

Based on the model scores, participants were successfully stratified into high, moderate, and low-risk tiers.

More results

The high-risk group exhibited a 4.2-fold increase in the incidence of adverse maternal and fetal outcomes compared to the low-risk group.

“

Conclusion · 1 of 2

The integration of angiogenic proteomics and clinical features via ensemble learning provides a high-performance and interpretable tool for the precision risk stratification of early-onset preeclampsia.

Conclusion · 2 of 2

This multi-dimensional approach facilitates early identification of high-risk individuals, potentially enabling more targeted clinical interventions and improved obstetric outcomes.

Integration of angiogenic pathway-related proteomics and clinical features via ensemble learning for precision risk stratification of early-onset preeclampsia.

Integration of angiogenic pathway-related proteomics and clinical features via ensemble learning for precision risk stratification of early-onset preeclampsia.

Related papers