We established that optimizing a machine learning model to predict in vitro fertilization (IVF) embryo implantation success is not an optimal strategy for the task of embryo ranking. The reason for this is “shortcut learning”, the model relies on the clinical factor as a proxy for implantation - hampering its ability to approximate the embryo quality.
Project led by Itay Elich.
Brief Twitter summary: https://x.com/AssafZaritsky/status/1831213767457227134
Comments