Artificial Intelligence in Obstetrics and Prenatal Medicine: Current Evidence, Clinical Validation and Implementation

Abstract Artificial intelligence (AI) is increasingly entering obstetrics and prenatal medicine, particularly in prenatal ultrasound, fetal echocardiography, fetal magnetic resonance imaging, genomic screening, risk prediction and clinical decision support. The aim of this review was to summarize current evidence, recent developments and implementation requirements for AI-based applications in prenatal care. A structured narrative review was performed with emphasis on peer-reviewed and clinically relevant literature published from 2021 to 2026, supplemented by landmark methodological, regulatory and reporting guidance. Priority was given to systematic reviews, diagnostic accuracy studies, external validation studies, prospective evaluations, randomized educational or clinical studies, regulatory documents and reporting standards relevant to AI in medicine. The strongest current evidence supports AI-assisted standard-plane recognition, image-quality assessment and fetal biometry, followed by emerging applications in fetal cardiac screening, blind-sweep ultrasound, fetal MRI segmentation, placental phenotyping, prediction of fetal growth restriction, preterm birth and preeclampsia, and machine learning-enhanced cfDNA analysis. Generative and multimodal AI systems may support report drafting, guideline retrieval and education, but remain insufficiently validated for autonomous clinical decision-making. AI has the potential to improve access, standardization, efficiency and quality assurance in prenatal medicine. However, most applications remain limited by retrospective design, enriched datasets, insufficient external validation, unclear calibration, limited fairness analyses and incomplete workflow evaluation. Clinical adoption should therefore follow a human-in-the-loop model with prospective validation, transparent reporting, regulatory oversight, bias monitoring and post-market surveillance. AI should augment rather than replace specialist expertise in obstetrics and fetal medicine.