Japan-Medical Image Database (J-MID): Medical Big Data Supporting Data Science

Area of Science:

• Medical Imaging Informatics
• Radiology Data Management
• Artificial Intelligence in Healthcare

Background:

• The increasing volume of radiological data, driven by advancements in imaging technologies and adherence to standards like Digital Imaging and Communications in Medicine (DICOM), poses challenges for diagnostic efficiency and quality.
• A limited number of radiologists in Japan face difficulties in processing this data surge, impacting diagnostic quality and efficiency.
• The Japan Radiological Society (JRS) launched the "Japan Safe Radiology" Initiative to improve safety, efficiency, and quality in radiological medicine through information and communication technology (ICT) utilization.

Purpose of the Study:

• To establish a national image diagnosis database to centralize medical resources and facilitate the efficient processing of radiological data.
• To leverage artificial intelligence (AI) and big data systems to integrate radiological medicine and improve diagnostic capabilities.
• To create an unparalleled repository of real-world radiological data for research and development.

Main Methods:

• The Japan Agency for Medical Research and Development (AMED) supported the development of the Japan Medical Image Database (J-MID) starting in 2017.
• J-MID systematically collects anonymized CT/MR images and diagnostic reports from 10 major Japanese university hospitals via the academic information network (SINET).
• Data is stored on a centralized cloud server, ensuring convenient access for researchers.

Main Results:

• As of April 2024, J-MID has amassed over 534 million images, representing 1.65 million cases.
• This makes J-MID the largest repository of real-world radiological data in Japan.
• The database facilitates the convenient utilization of extensive medical imaging data by researchers.

Conclusions:

• The establishment and growth of J-MID represent a significant advancement in managing and utilizing large-scale radiological data in Japan.
• This initiative supports the "Japan Safe Radiology" Initiative by providing a robust platform for AI-driven advancements and improved diagnostic efficiency.
• J-MID serves as a crucial resource for advancing medical imaging research and enhancing the quality of radiological diagnoses nationwide.