phenotyping

Unsupervised machine learning identified three distinct preeclampsia phenotypes linked to timing of delivery, placental dysfunction, fetal growth, and complications, suggesting a path toward more personalized risk assessment and management.