Early categorisation of bovine embryos to boost IVF success

Background and challenges

Currently, in cattle, the success rate of IVF and embryo transfer does not exceed 30% of viable births. The selection of embryos (from oocytes collected in vivo or post mortem and then fertilised) is based on a classification at D7 after fertilisation. One of the keys to increasing IVF performance is to optimise this selection as early as possible.

Goals

The objective of the BovMovie2Pred consortium is to propose solutions to assist in the selection of bovine embryos in order to significantly increase the percentage of viable births from in vitro produced embryos.

The aim is to optimise the selection of embryos as early as possible by exploiting their morphokinetic history, from fertilisation to the day of transfer. This history is traced from annotated videos. However, expert annotations of videos have the double disadvantage of being laborious to carry out and having a subjective element.

In order to overcome these constraints, the BovMovie2Pred consortium proposes to organise one or more data challenges within the framework of the RAMP (Rapid Analytics and Model Prototyping) platform of the DATA-IA Convergences Institute. These challenges will bring together the skills of experts on AI issues as well as those of students or PhD students in this field. The expertise of the consortium, coupled with existing annotation work, will make it possible at the end of the project to provide researchers in developmental biology with a classification methodology requiring as little video annotation as possible.

Contact

Partnerships

INRAE participants

Partners

Publications

Journal article

Articles

Hachani Y., Bouthemy P., Fromont E., Ruffini S., Laffont L., de Paula Reis A. 2025. Prediction of cell stages and cleavage durations of IVP bovine embryos with a deep learning model. Proceedings of the 41th Scientific meeting of the Association of Embryo Technology in Europe (AETE), Septembre 2025, Cork, Irlande. Publié dans: Animal Reproduction, 2025; 22(3). (abstract) pas dans HAL

Hachani Y., Bouthemy P., Fromont E., Ruffini S, Laffont L., et al.. Supervised contrastive learning for cell stage classification of animal embryos. Scientific Reports, 2026, ⟨10.1038/s41598-026-39214-y⟩. ⟨hal-04937720v3⟩

Conferences papers

• Hachini Y., Bouthemy P., Fromont E., Ruffini S., Laffont L., de Paula Reis, A. 2024. Early prediction of the transferability of bovine embryos from videomicroscopy. Proccedings of the 2024 IEEE International Conference on Image Processing (ICIP), Octobre 2024, Abu Dhabi, Emirats Arabes Unis. DOI: 10.1109/ICIP51287.2024.10647901
• Hachini Y., Bouthemy P., Fromont E., Ruffini S., Laffont L., de Paula Reis. 2024. Prédiction précoce de la transférabilité d'embryons bovins par vidéomicroscopie. Congrès RFIAP 2024 - Reconnaissance des Formes, Image, Apprentissage et Perception, AFRIF (Association Française pour la Reconnaissance et l’Interprétation des Formes), Juillet 2024, Lille, France.
•   ​​​​​​Hachini Y., Bouthemy P., Fromont E., Ruffini S., Laffont L., de Paula Reis. 2024. Early prediction of the transferability of bovine embryos from videomicroscopy. Colloque IABM 2024 - Colloque Français d’Intelligence Artificielle en Imagerie Biomédicale, Mars 2024, Grenoble, France.