Deep Learning Models: Their Relationship with Embryonic Euploidies and Reproductive Outcomes

Background: Embryo selection in in vitro fertilization (IVF) aims to prioritize embryos with the highest reproductive potential. While preimplantation genetic testing for aneuploidy (PGT-A) remains the gold standard for identifying euploid embryos, it is invasive and not universally applicable. Deep learning (DL)-based models, such as the intelligent data analysis (iDA) score, have emerged as non-invasive alternatives for embryo assessment. This review critically evaluates the relationship between iDAScore (versions 1.0 and 2.0), embryo euploidy, and clinical outcomes, including live birth and miscarriage rates. Methods: A narrative review was performed using PubMed and Google Scholar, covering studies published from January 2020 to May 2025. The search included terms such as "iDAScore," "deep learning," "euploidy," and "live birth." Only English-language full-text studies assessing the predictive performance of iDAScore relative to chromosomal status or reproductive outcomes were included. Results: Six retrospective studies met the inclusion criteria. All reported a statistically significant association between higher iDAScore values and embryo euploidy. AUC values for euploidy prediction ranged from 0.60 to 0.68. In several studies, iDAScore was also positively associated with live birth rates and negatively with miscarriage rates. However, the predictive accuracy was moderate when restricted to euploid embryo cohorts, indicating that iDAScore may be more effective in broader populations where chromosomal status is unknown. Conclusions: iDAScore represents a promising adjunct to traditional embryo assessment. Although it cannot replace PGT-A, it may aid in embryo prioritization when genetic testing is not feasible. Larger prospective studies are warranted to further validate its clinical utility.