Using Automated Cell Tracking to Improve Embryo Viability Prediction for In Vitro Fertilization

Nearly one in six couples in the United States experience infertility, many of whom turn to in vitro fertilization (IVF) to have children. However, only 20-25% of IVF cycles result in a live birth because clinicians can only assess embryo viability based on visual morphology, which has been shown to be only weakly correlated to viability [1]. Because of this low success rate, it is common to transfer several embryos back to the mother at once in hopes that at least one will survive. This causes a high rate of multiple births (~30%), which are associated with increased health complications for both mothers and children [2]. A more accurate predictor of embryo viability is needed so that clinicians can accurately choose a single embryo to transfer back to the mother, thus decreasing the rate of multiple births while increasing the overall IVF success rate.