Advancing embryo selection in artificial intelligence-assisted reproductive technologies: A systematic review
Artificial Intelligence in Health · 2025-05-02 · 2 (3) · 1
DOI: 10.36922/aih.7170 ↗Abstract
For couples encountering infertility challenges, assisted reproductive technologies (ARTs) offer a path to parenthood. ART procedures, such as in vitro fertilization (IVF), intracytoplasmic sperm injection (ICSI), and embryo implantation, involve the handling of sperm or embryos outside the body. However, the success of ART depends on the accurate selection of viable embryos. Artificial intelligence (AI) is a promising tool with the potential to revolutionize these procedures. This review explores the transformative potential of AI in ART, providing valuable insights into enhanced embryo selection and unlocking new possibilities for the field. Four electronic databases were systematically searched under the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. From an initial pool of 914 papers, 30 studies were selected for further evaluation. While noting the limitations inherent in the existing body of research, this review offers a broad analysis of AI’s transformative role in embryo selection. It highlights the significant potential of AI to enhance precision, consistency, and efficiency in ART. This review also emphasizes the importance of addressing technical, ethical, and regulatory aspects to ensure responsible and effective integration of these technologies. The findings indicate that AI-based models, such as the iDAScore v2.0, have demonstrated promising results in accurately predicting embryo viability and evaluating the effects of maternal age on embryo viability. Specifically, Bayesian network modeling, with an accuracy rate of 91.3%, aims to optimize IVF and ICSI procedures. In summary, AI stands at the forefront of innovation in ART, offering new hope through more accurate and efficient embryo selection.
EUVIMED