Artificial intelligence (AI) integrated with time-lapse (TL) imaging for embryo selection significantly minimizes subjectivity and workload in traditional methods, marking a pivotal advancement in the realm of assisted reproductive technology. This review comprehensively summarizes the representative studies conducted in recent years within these domains and delves into the application of AI combined with TL imaging for embryo selection from three perspectives: data selection, model selection, and model evaluation. While AI research has yet to fully achieve automated embryo selection, it has already commenced assisting embryologists in annotating and selecting embryos to a considerable extent, thereby reducing subjective discrepancies and easing the workload. Existing software for oocyte quality scoring, embryo ploidy prediction, and transfer outcome forecasting has exhibited promising performance in studies. However, there remains an ongoing need for the development of high-quality datasets and the conduct of prospective, multicenter studies with large sample sizes across diverse devices. The evolution of AI combined with TL imaging in the field of embryo selection merits heightened attention from both clinicians and embryologists.

Artificial intelligence (AI) integrated with time-lapse (TL) imaging for embryo selection significantly minimizes subjectivity and workload in traditional methods, marking a pivotal advancement in the realm of assisted reproductive technology. This review comprehensively summarizes the representative studies conducted in recent years within these domains and delves into the application of AI combined with TL imaging for embryo selection from three perspectives: data selection, model selection, and model evaluation. While AI research has yet to fully achieve automated embryo selection, it has already commenced assisting embryologists in annotating and selecting embryos to a considerable extent, thereby reducing subjective discrepancies and easing the workload. Existing software for oocyte quality scoring, embryo ploidy prediction, and transfer outcome forecasting has exhibited promising performance in studies. However, there remains an ongoing need for the development of high-quality datasets and the conduct of prospective, multicenter studies with large sample sizes across diverse devices. The evolution of AI combined with TL imaging in the field of embryo selection merits heightened attention from both clinicians and embryologists.