Retinal Imaging and Artificial Intelligence for the Screening and Risk Stratification of Preeclampsia: A Narrative Review

Preeclampsia is a leading cause of maternal and perinatal morbidity and mortality worldwide. However, current screening strategies remain limited by low positive predictive values, thus highlighting the urgent need to develop novel and non-invasive approaches. The retina provides a uniquely accessible window into systemic microvascular health, as retinal vessels share embryological origin, autoregulatory mechanisms, and susceptibility to endothelial dysfunction with other microvascular beds that are central to the pathophysiology of preeclampsia. Advances in fundus photography, optical coherence tomography, and optical coherence tomography angiography have enabled the quantitative assessment of retinal vascular changes that are detectable before the onset of clinical disease and correlate with disease severity and adverse maternal-fetal outcomes. When integrated with clinical parameters, artificial intelligence (AI)-based predictive models have demonstrated promising performance for early preeclampsia risk stratification, outperforming conventional clinical models. This narrative review synthesizes and critically appraises the current evidence on retinal imaging and AI-based models for preeclampsia screening, identifies key research gaps, and discusses the challenges for clinical translation.