A joint study with Keio University and Tokyo University has been published in Sci. Rep.

Professor Masato Ikawa from the Dept. of Experimental Genome Research at our institute, in collaboration with Keio University and the University of Tokyo, has developed a new method using deep learning—an AI technology—to identify the stages of sperm development in mice.

This method analyzes stained microscope images of mouse testes and automatically classifies the seminiferous tubules, where sperm are produced, into 12 distinct stages. The system demonstrated extremely high accuracy, achieving a remarkable 98.33% prediction rate.

This technology is expected to contribute to more precise and efficient diagnosis and research in reproductive and fertility medicine in the future.

Related Links

• Details of the Research Results (Keio University Press Release, ResOU Osaka University)
• Publication in Scientific Reports
• Department of Experimental Genome Research​ Details