AI and Tokushukai Medical Big Data

The Tokushukai Group operates approximately 340 facilities nationwide, including 71 hospitals, and manages medical data on more than 13 million patients in the Tokushukai Medical Database (TMD).

Artificial intelligence (AI), which has attracted attention for its predictive capabilities, requires well-organised big data suitable for AI development. The Group can collect structured, up-to-date medical data from all its facilities, providing an environment for advanced AI development and research, including research on COVID-19. We use TMD to undertake a range of AI development and collaborative research projects.

Responding to Digital Transformation

Digital transformation (DX) is sweeping through businesses around the world.

Practices once taken for granted are changing significantly. Just as bank counter services are disappearing, regular in-hospital consultations may also disappear in the near future. Organisations are being called on to innovate using AI.

One example of innovation in healthcare is the rapid development of clinical decision support systems (CDSS) in the United States, some of which are already in practical use.

CDSS use computers and AI to help healthcare professionals make treatment and diagnostic decisions and provide better care. They are among the systems the Tokushukai Group will need in the future. As one form of CDSS, we introduced the UK's National Early Warning Score (NEWS) at 28 Group hospitals to help prevent unexpected in-hospital cardiac arrest. To provide better care, the Group has begun developing AI with greater predictive accuracy than NEWS. We will continue pursuing new innovations, including CDSS.

Our Approach to AI Development

Machine learning, a form of AI, learns knowledge and predictive patterns from big data. It is a useful technology that provides people with appropriate predictive information and advice while rapidly processing quantities of data beyond human capacity.

The Information System Department works with TMD and medical facilities, including doctors, pharmacists, nurses and systems engineers, to use Group resources for AI development. Projects include predicting severe COVID-19 and cardiac arrest, and suggesting medications for patients with diabetes. When collaborating with doctors and other researchers, we begin by discussing the relevant literature and the data to extract, such as CT images and blood test results. Our engineers then extract structured data from TMD and propose suitable approaches using statistical analysis, machine learning and other methods. We share our research through conference presentations, listed below, and continue our steady work towards new innovations.

Conference Presentations

Please see our list of presentations for details.