JBRA Assisted Reproduction 2025;29(Suppl.2 SBRA 2025):121
Poster Presentation
29^th Annual Congress of the SBRA. São Paulo/SP - Brazil, 2025
doi: 10.5935/1518-0557.20263755

P-109. From Embryology to Artificial Intelligence: A New Era in Decision Making in Assisted Reproduction

Andrea Mesquita Lima¹, Gleicyane Sousa Santos Alam¹, Sebastião Evangelista Torquato Filho¹, Tulius Augustus Ferreira Freitas¹, Ellayne Cavalcanti Queiroz¹, Eduardo Gomes Sá¹, Renata Reis Pimentel Castro¹

¹ Bios Centro de Medicina Reprodutiva – Fortaleza -CE - Brasil

Objective: Assisted reproduction faces the constant challenge of optimizing clinical outcomes based on multiple biological and laboratory variables. The advancement of artificial intelligence (AI), especially machine learning, offers a promising approach to support more accurate and reproducible decisions, especially in embryo selection. The objective of this work is to present the application of AI algorithms in critical processes of in vitro fertilization (IVF), comparing their results with conventional methods and evaluating their impact on the quality and predictability of clinical outcomes.
Methods: Retrospective data from IVF cycles performed in an accredited clinic over a period of 24 months were analyzed. The study compared the performance of embryo selection performed by experienced embryologists with that performed by an AI algorithm trained with morphokinetic data and clinical outcomes. The analysis included implantation, clinical pregnancy and live birth rates. Furthermore, the integration of AI into the clinic's quality management system was evaluated, observing laboratory indicators before and after its implementation.
Results: AI showed higher accuracy in predicting implantation than embryologists in 8.4% of cases (p<0.05), with greater inter-cycle consistency. The implantation rate increased by 12% in the AI-supported group, especially in patients over 38 years of age. The analysis integrated with the quality management system demonstrated better traceability of decisions and reduction of subjective variations. The ethical challenges of adopting AI, the need for multicenter validation and the role of the embryologist as information curator are also discussed.
Conclusion: The incorporation of artificial intelligence in assisted reproduction has the potential to transform clinical practice, offering greater precision, predictability and standardization. Far from replacing the embryologist, AI presents itself as a powerful decision-support tool, standing out as a strategic ally in environments with a strong culture of quality and innovation.