July 2, 2025
Keio University
Research Institute for Microbial Diseases, Osaka University
Graduate School of Agricultural and Life Sciences / Faculty of Agriculture, The University of Tokyo
A research group including Yudai Tokuoka, a researcher; Shun Morikura, a Project Assistant Professor／Project Research Associate／Project Instructor; and Professor Kei Funahashi from the Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University; Assistant Professor Ken Endo from the Center for Experimental Animals, Tokyo Medical and Dental University (now Tokyo University of Science) (currently Assistant Professor at the Graduate School of Agricultural and Life Sciences, The University of Tokyo); and Yuki Hiradate, a then Project Assistant Professor／Project Research Associate／Project Instructor, and Professor Masahito Ikawa from the Research Institute for Microbial Diseases, Osaka University, has successfully developed an algorithm using deep learning to accurately identify the 12 stages of seminiferous tubules from bright-field microscope images of histologically stained mouse tissue. They further demonstrated that the classification accuracy of the stage prediction is extremely high, at 98.33%, when allowing for a prediction error of ±1 stage. This method is expected to contribute to the fields of assisted reproductive technology and infertility treatment as a new foundational technology for the automatic and quantitative evaluation of seminiferous tubule stages.
The results of this research were published as an advance online publication on the website of the academic journal Scientific Reports on July 1 (UK time).
The full press release is available below.