Abstract Background Artificial Intelligence (AI) has become one of the most actively developing instruments in obstetric ultrasound, bringing new opportunities to better fetal evaluation, decrease operator dependence, and to improve the accuracy of diagnosis. The need to provide high-quality prenatal imaging, together with unequal access to adequately trained sonographers across different healthcare settings and regions, has driven the adoption of AI-driven solutions. Main body Recent studies show that AI can support automated fetal biometry, gestational-age estimation, standard-plane detection, and early prediction of congenital abnormalities. In clinical trials, deep learning systems have demonstrated the ability to achieve diagnostic performance comparable to that of experienced sonographers, particularly in routine measurements and image-quality optimization. Nevertheless, the generalizability of these models remains challenged by variability in ultrasound devices, population characteristics, and image-acquisition methodologies. Clinical implementation is further complicated by ethical concerns, including data privacy, algorithmic bias, and transparency. This article is a summary of current peer-reviewed research, and it outlines the potentials and shortcomings of AI-based obstetric imaging. The major themes include existing applications, diagnostic accuracy, implementation challenges, and emerging directions such as federated learning, explainable AI, and point-of-care integration. Conclusion Obstetric ultrasound AI has a bright future of improving prenatal care, and its widespread implementation needs strict validation, moral protection, and implementation in clinical practices. As more research is done and AI is applied attentively, the algorithm can potentially enable more accessible, accurate and more equitable maternal–fetal health outcomes.
Mahmoud et al. (Thu,) studied this question.