Artificial Intelligence in Recurrent Pregnancy Loss: Current Evidence, Limitations, and Future Directions

BACKGROUND: Despite significant advances in genetics, immunology, and endometrial research, the underlying cause of nearly half of recurrent pregnancy loss (RPL) cases remains unknown. This highlights the limitations of conventional diagnostic approaches and underscores the need for methods that can detect complex, subtle biological patterns. OBJECTIVES: To summarize and critically assess how artificial intelligence (AI) is changing our knowledge of, ability to predict, and future therapeutic management of RPL, with a focus on machine learning (ML) approaches that identify latent biological pathways and multifactorial contributors to pregnancy loss. METHODS: This narrative review summarizes contemporary research on AI applications in reproductive medicine. Research using imaging, proteomic, genomic, clinical, and multi-omics information to create predictive or mechanistic models associated with RPL provided evidence. RESULTS: AI-based approaches are increasingly demonstrating the ability to detect complex interactions among environmental, immunological, biochemical, and genetic factors associated with RPL. ML and deep learning (DL) models enhance prognostic accuracy, identify novel candidate biomarkers, and provide insights into the systemic and molecular mechanisms underlying pregnancy loss. Integrating heterogeneous data through AI supports the development of personalized reproductive profiles and can improve prediction and counseling. CONCLUSIONS: AI has the potential to improve both personalized prediction and mechanistic understanding of RPL. However, clinical translation is currently hampered by a number of important issues, including small and diverse datasets, conflicting diagnostic definitions, limited external validation, and a lack of prospective clinical trials. To responsibly integrate AI tools into reproductive care, these limitations must be addressed.