New Publication: Artificial intelligence models developed by Instituto Bernabeu reliably predict reproductive success

Instituto Bernabeu has published a new study in the scientific journal Reproductive Biology and Endocrinology that represents a significant advance in the use of artificial intelligence applied to reproductive medicine. The work, led by Dr. Jorge Ten, Director of our Embryology Unit, analyzes and compares different machine-learning algorithms to determine which embryonic development parameters best predict implantation and the birth of a healthy baby in treatments using oocyte donation.

The study is based on the analysis of 378 donor-egg embryos transferred in single-embryo cycles at four Instituto Bernabeu centers, all with known clinical outcomes for implantation and live birth. The researchers evaluated the performance of ten artificial intelligence models and concluded that the Random Forest algorithm is the most effective for predicting implantation, while AdaBoost delivers the best performance for predicting births, with accuracy approaching 80%.

One of the most relevant findings of the study is that the interval from oocyte fertilization to the onset of blastocyst hatching—when the embryo “breaks” through its natural envelope to enable implantation—is the most decisive parameter for anticipating treatment success.

Dr. Ten explains: “These results strengthen the usefulness of artificial intelligence as a decision-support tool for embryologists and allow us to identify with greater precision which embryos have the highest potential to result in a birth.” This advance, part of our Artificial Intelligence Unit, reinforces Instituto Bernabeu’s commitment to technological innovation and applied research to improve clinical outcomes and the safety of our patients.

Dr. Jorge Ten, Director of the Embryology Unit at Instituto Bernabeu