November 4, 2022
Keio University
Kindai University
Institute of Industrial Science, The University of Tokyo
A group of researchers, including Yudai Tokuoka from the Keio Leading-edge Laboratory of Science and Technology (KLL), Professor Akira Funahashi and Senior Assistant Professor Takahiro Yamada from the Faculty of Science and Technology at Keio University, Professor Kazuo Yamagata from the Faculty of Biology-Oriented Science and Technology at Kindai University, and Associate Professor Tetsuya Kobayashi from the Institute of Industrial Science at The University of Tokyo, has successfully developed an AI (NVAN) that uses deep learning to accurately predict live birth from data continuously captured from the cell division of mouse embryos. The classification accuracy of NVAN's live birth prediction is an astonishing 83.87%, surpassing the world's most advanced machine learning methods and visual inspections by experienced embryologists. This method is expected to contribute to human assisted reproductive technology and livestock production as a new foundational technology for embryo evaluation in in vitro fertilization.
The results of this research were published as an online early version on the website of the academic journal Artificial Intelligence in Medicine on November 2, 2022 (UK time).
Please see below for the full press release.