Evaluation of the Use of Artificial Intelligence (AiVF) in Embryo Selection in Assisted Reproduction for Embryos Cultured to the Fifth Day of Development

Embryo selection is one of the major challenges in Assisted Reproductive Technology (ART) in order to continuously improve clinical pregnancy rates. Morphological evaluation of embryos is the standard method used to assess embryo quality and to select the embryo to be transferred. However, this observational method has limitations: it can be subjective and may lead to inter- and intra-observer variability. To increase pregnancy rates, the transfer of two embryos may be proposed, which can result in high-risk multiple pregnancies. Despite the introduction of time-lapse systems and embryo selection algorithms aimed at increasing live birth rates, outcomes in ART remain suboptimal and can still be improved.

The use of artificial intelligence (AI) is expanding in the medical field and is increasingly being applied in ART laboratories to optimize outcomes while avoiding multiple pregnancies. EMA™ is a secure artificial intelligence software platform developed by the company AiVF. The use of this software could allow a standardized and reproducible evaluation of embryo development up to the fifth day of embryonic culture. Through the application of an algorithm, the software may enable the selection of the embryo with the highest potential for achieving pregnancy.

The primary objective of this study is to evaluate the use of the EMA™ artificial intelligence system in embryo selection for embryos cultured up to day 5, in comparison with the morphokinetic criteria currently used in the laboratory for the first fresh transfer of a day-5 embryo.

The secondary objectives are to assess whether artificial intelligence (1) reduces the time to achieve pregnancy (defined as pregnancy achieved or not after the first transfer of a day-5 embryo) and (2) affects the miscarriage rate among patients treated in the ART CECOS unit. A medico-economic analysis will be conducted to evaluate the number of new patients undergoing an additional treatment attempt and to determine whether there is a correlated reduction in time to pregnancy by assessing the number of pregnancies achieved after the first transfer in patients already under treatment. A cost-effectiveness analysis will also be performed, comparing the cost-effectiveness ratios between the two study arms.