Fujifilm Launches "SYNAPSE Creative Space" Service in Japan

TOKYO, April 1, 2024 – FUJIFILM Corporation (President and CEO, Representative Director: Teiichi Goto) announces that it will launch the "SYNAPSE Creative Space," a service to support the development of AI technologies for image diagnosis at medical institutions and research organizations in Japan, starting April 12. The "SYNAPSE Creative Space" is a cloud service that enables doctors and researchers to develop their own AI technologies to support image diagnosis. It does not require any special knowledge such as programming^*1. The service is based on an AI technology development research platform system jointly developed by Fujifilm and the National Cancer Center Japan.

There are three major hurdles to developing AI technology to support image diagnosis: first, it requires a high-performance server and development environment; second, it requires advanced engineering skills such as programming to design the learning model and execute the development process; and third, it requires many hours and effort to annotate or generate data for the AI to learn. In particular, the annotation process requires several minutes to an hour per case, and the work must be performed on hundreds or thousands of datasets, making it a major burden in the AI technology development process.

The "SYNAPSE Creative Space" is a service that enables a series of development processes, such as project management, annotation, learning, and AI technology testing, in the development of AI technology to support image diagnosis to be performed all at once in the cloud. Since the development environment is ready on the cloud, it can be easily used by simply preparing a PC at hand. In addition, since multiple learning models for medical images developed jointly by Fujifilm and the National Cancer Center Japan can be used, doctors and researchers can develop AI technology without having to design their own learning models from scratch. Furthermore, the annotation tool for this service has a screen design similar to that of Fujifilm’s AI platform available in Japan and used in clinical settings, allowing efficient and intuitive reference and annotation of images with a feel similar to that of the diagnostic imaging environment.

The "SYNAPSE Creative Space" has evolved to be even easier to use, including enhanced annotation support features that reflect requests received during trials conducted at more than 40 medical and research institutions in Japan since April 2022. It strongly supports the development of AI technology to support image diagnosis. It can also be used not only as a support tool for research and development, but also as a teaching tool for hands-on education of medical professionals and medical students. It allows students to experience a part of AI technology development without engineering knowledge, thus helping them understand the basic principles and mechanisms of AI technology, which is increasingly used in clinical settings, and enhancing AI literacy.

It is expected that the widespread use of the "SYNAPSE Creative Space" will promote the development of AI technologies to support image diagnosis for various diseases, including rare diseases. Fujifilm has already conducted joint research with the National Cancer Center Japan and Nagoya City University to develop two AI technologies^*2. The service has also been piloted at six medical schools, graduate schools and medical colleges to support AI training for more than 300 medical students. By providing this service, Fujifilm aims to support the development of AI technology while also focusing on social implementation and educational support.

• *1 No specialized knowledge is required when using the learning model provided by this service. Specialized knowledge and skills are necessary when designing a new learning model itself.
• *2 Fujifilm and the National Cancer Center Japan have jointly developed AI technology to extract areas of suspected glioma from MRI images at a precise level. Fujifilm and Nagoya City University have jointly developed AI technology to extract each region of the cerebrospinal fluid cavity from MRI images.