Trial emulation for validating the clinical efficacy of a foundational AI model in embryo selection

Autores de IIS La Fe

• Laura Arnal Benedicto

  Autor

• 

  Marcos Meseguer Escrivá

  Autor

Participantes ajenos a IIS La Fe

• Rajendran S.
• Malmsten J.E.
• Rosenwaks Z.
• Zaninovic N.
• Hajirasouliha I.

Grupos

• Biomarcadores, Medicina Genómica, Estadística y Análisis masivo de datos en Reproducción humana asistida

Abstract

AI models for embryo selection often rely on correlational metrics that ignore clinical confounders. We present a target trial emulation framework to approximate causal effects for a foundational AI model, FEMI, for non-invasive embryo assessment using multi-center trial emulation (n = 4674). Propensity score matching established a robust association between FEMI-Ploidy and implantation failure showing an average treatment effect (ATE) of -0.131 (95% CI [-0.196, -0.066]) in the development cohort and -0.157 (95% CI [-0.254, -0.054]) in the external cohort. Comparative efficacy using S-Learner models demonstrated that a high-risk FEMI score carried a significantly stronger individual treatment effect (ITE) penalty on implantation compared to other scoring mechanisms (p < 0.0001). This superiority persisted after adjusting for maternal age, suggesting FEMI captures unique biological features. This causal framework establishes a rigorous standard for AI validation in IVF, providing the necessary pre-clinical justification for prospective randomized controlled trials. © The Author(s) 2026.

Keywords

• Average treatment effects; Biological features; Clinical efficacy; Confounder; Emulation framework; Learner modeling; Maternal age; Propensity score matching; Robust association; Treatment effects; adult; aneuploidy; Article; artificial intelligence; blastocyst; calibration; causal attribution; chromosome analysis; clinical effectiveness; clinical outcome; cohort analysis; comparative effectiveness; controlled study; embryo selection; female; human; in vitro fertilization; infertility therapy; live birth; major clinical study; maternal age; multicenter study (topic); nidation; patient counseling; prediction algorithm; pregnancy outcome; propensity score; retrospective study; Spain; treatment failure; treatment response; utility value; validation study; Artificial intelligence