Automated Evaluation of Human Embryo Blastulation and Implantation Potential using Deep‐Learning. (1st July 2020)

• Record Type: Journal Article
• Authors: Kan-Tor, Yoav; Zabari, Nir; Erlich, Ity; Szeskin, Adi; Amitai, Tamar; Richter, Dganit; Or, Yuval; Shoham, Zeev; Hurwitz, Arye; Har-Vardi, Iris; Gavish, Matan; Ben-Meir, Assaf; Buxboim, Amnon
• Is Part Of: Advanced intelligent systems. Volume 2:Number 10(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-07-01
• Subjects: automated embryo classification -- deep learning -- embryo transfers -- in vitro fertilization; Artificial intelligence -- Periodicals; Robotics -- Periodicals; Control theory -- Periodicals; 006.3
• DOI: 10.1002/aisy.202000080
• Languages: English
• ISSNs: 2640-4567

Abstract

Abstract : In in vitro fertilization (IVF) treatments, early identification of embryos with high implantation potential is required for shortening time to pregnancy while avoiding clinical complications to the newborn and the mother caused by multiple pregnancies. Current classification tools are based on morphological and morphokinetic parameters that are manually annotated using time‐lapse video files. However, manual annotation introduces interobserver and intraobserver variability and provides a discrete representation of preimplantation development while ignoring dynamic features that are associated with embryo quality. A fully automated and standardized classifiers are developed by training deep neural networks directly on the raw video files of >6200 blastulation‐labeled and >5500 implantation‐labeled embryos. Prediction of embryo implantation is more accurate than the current state‐of‐the‐art morphokientic classifier. Embryo classification improves with video length where the most predictive images show only partial association with morphological features. Deep learning substitute to human evaluation of embryo developmental competence thus contributes to implementing single embryo transfer methodology. Abstract : Human reproduction is inefficient. A significant fraction of the fertilized oocytes lacks the capacity to advance through preimplantation embryo development and to implant within the uterus. Using embryo time‐lapse images from IVF clinics, deep neural networks are trained to assess the potential to reach blastulation and generate pregnancy. The classifiers are automated, standardized, and improve accuracy over manually annotated algorithms.