This team performs AI-based editing and analysis of surgical videos. We aim to create shortened versions that extract key points from videos ranging from several hours to over ten hours long, and to develop systems for automatically generating surgical records. By reducing the effort required for video management, we contribute to securing more time for research.

We connect local LLMs to electronic medical records to safely process sensitive information, while simultaneously utilizing cloud LLMs to support tasks such as slide creation and translation. We optimize the environment according to the nature of the information, supporting the efficiency of administrative work in medical settings.

Advanced data science skills and deep knowledge of medical informatics are essential for developing “actually useful AI.” We lead the appropriate handling of data and analysis methods in each project, supporting the foundation for high-quality medical AI development.