Integrating Artificial Intelligence into Perinatal Care Pathways: An Umbrella Review of Applications, Outcomes, and Equity

Abstract

C (Comparator): Standard clinical practice without AI/ML assistance (e.g., clinician-only decision making, conventional image interpretation, manual vital-sign monitoring).O (Outcomes): Predictive performance metrics (e.g., accuracy, sensitivity, AUC, Dice coefficient), clinical usability (e.g., time-to-decision, workflow integration), and downstream patient-centered outcomes (e.g., reduction in adverse events, length of stay, neonatal morbidity/mortality). S (Study Designs): Systematic reviews, scoping reviews, narrative reviews, and meta-analysesTo synthesize existing review-level evidence on AI/ ML applications spanning reproductive, prenatal, postpartum, and neonatal care-evaluating (1) model predictive performance (e.g., accuracy, AUC, Dice), (2) clinical usability and integration in existing workflows, and (3) reported impacts on patient outcomes-by categorizing findings into four domains (reproductive/early childhood, pregnancy, postpartum, neonatal).The ultimate aim is to identify strengths, limitations, gaps, and future research priorities for AI/ML adoption in perinatal healthcare. RationaleAI and ML have proliferated in maternal, f e t a l , a n d n e o n a t a l d o m a i n s , p ro m i s i n g improvements in early diagnosis, risk stratification, INPLASY 1

Ähnliche Arbeiten