Applications of artificial intelligence in obstetrics

review article

Abstract

Introduction: Artificial intelligence (AI) has emerged as a transformative technology in medicine, particularly in obstetrics and maternal–fetal medicine. By enabling automated analysis of large volumes of clinical and imaging data, AI offers new opportunities to improve screening, prediction, diagnosis, and clinical decision support. Despite the rapid growth of AI applications, concerns remain regarding methodological standardization, external validation, and real-world clinical applicability. Objective: To describe and critically analyze the main applications of artificial intelligence in obstetrics and maternal–fetal medicine, focusing on diagnostic and predictive performance, methodological limitations, and potential impact on clinical practice. Methods: A literature review was conducted using the PubMed database, including studies published between 2021 and 2025. Original research articles, systematic reviews, and meta-analyses addressing practical applications of AI in obstetrics and reporting algorithm performance metrics were included. The selected studies were categorized into four main domains: gestational age and fetal weight estimation, fetal and neonatal neurological assessment, prediction of hypertensive disorders of pregnancy, and prenatal screening for structural and genetic anomalies. Results: Twelve studies met the inclusion criteria. AI-based systems demonstrated performance comparable to that of experienced clinicians in gestational age estimation, particularly in low-resource settings. In fetal and neonatal neurological assessment, deep learning models showed high accuracy in detecting cerebral lesions and enabled dynamic evaluation of fetal brain activity through automated facial expression recognition. For the prediction of hypertensive disorders of pregnancy, AI models—especially those based on placental texture analysis—showed promising results, although with considerable methodological heterogeneity. AI-assisted screening for structural and genetic anomalies also achieved robust diagnostic performance, reducing operator dependency. Conclusion: Current evidence suggests that artificial intelligence has the potential to enhance diagnostic accuracy, reduce interobserver variability, and improve efficiency in obstetric care. However, challenges related to model interpretability, external validation, generalizability, and safe clinical integration remain. AI should be regarded as a complementary tool to clinical judgment rather than a replacement, and its successful implementation requires evidence-based guidelines and adequate training of healthcare professionals.