External validation of a time lapse prediction model

This external validation study failed to replicate the performance of the time-lapse published prediction model, but it should inspire each center to build its own prediction model.

Thomas Fréour, Pharm.D., Ph.D., Nicolas Le Fleuter, Pharm.D., M.Sc., Jenna Lammers, M.Sc., Carole Splingart, Pharm.D., M.Sc., Arnaud Reignier, M.Sc., Paul Barrière, M.D., Ph.D.

Volume 103, Issue 4, Pages 917-922


To study the performance of a previously published implantation prediction model based on morphokinetics in a different setting, in an unselected population and with various embryo transfer strategies.

Retrospective monocentric study.

University-based assisted reproduction technology (ART) center.

450 unselected couples undergoing intracytoplasmic sperm injection (ICSI) cycle with embryo culture in the EmbryoScope (Unisense Fertilitech), corresponding to 528 embryos with known implantation.


Main Outcome Measure(s):
Implantation rates (IR) in embryo categories defined by the model in the overall population and in subgroups according to the day of embryo transfer.

The distribution of IR among detailed morphokinetic categories in the overall population and in subgroups according to the day of embryo transfer was more heterogeneous than expected according to the published model. The distribution corresponded better to the original when a simplified version of the model was used, although it worked better in the cleavage-stage group than in the blastocyst-stage group.

This study was unsuccessful in replicating the sensitivity of the previously published model for predicting implantation rate of embryos ranked according to morphokinetic categories. Further work is required to assess the utility of the model for embryo selection. Each team using time-lapse technology should build a center-specific prediction model based on its own data and transfer policy.

  • msamplaski

    Time-lapse technology is emerging as a commonly utilized tool in ART. However, its role and clinical protocols are still being determined. These results are different from the prior published literature, but they will likely contribute to the growing body of literature on the role of time-lapse technology. My primary question is: have the various groups who develop these devices demonstrated that there is consistency between devices? Are we comparing a Tercel to a Tesla?

Translate »