GIST launches W11b medical AI project to detect and treat cancer, brain diseases

South Korea is pushing to develop a next-generation medical AI platform that unifies three distinct systems: a "scientist AI" that identifies the causes of disease, a "clinician AI" that supports diagnosis and treatment, and a "personal physician AI" that monitors health in everyday life.

The Gwangju Institute of Science and Technology announced Tuesday it had been selected for an interdisciplinary project under the Ministry of Science and ICT and the Institute for Information and Communications Technology Planning and Evaluation's 2026 AI Top-Tier Early-Career Researcher Support Program.

The project will receive 11 billion won ($8.17 million) in government research and development funding over five and a half years through 2031. GIST will lead the project, with Sungkyunkwan University participating as a joint research and development institution. Lee Hyun-ju, a professor in GIST's Department of AI, will serve as principal investigator.

At the heart of the effort is the development of core technology called MIND — Medical INtelligence for Discovery, Diagnosis, and Daily care — which aims to connect disease detection, clinical decision-making and everyday health management in a single system.

The research team is developing three distinct AI components. The scientist AI will analyze omics data — including genomics and protein data — alongside medical imaging and single-cell data to uncover the mechanisms behind cancer and brain diseases, and to identify new biomarkers and therapeutic targets.

The clinician AI is designed to operate reliably across different hospitals, patients, medical devices and testing methods, predicting diagnoses, prognoses and treatment responses. Rather than simply presenting results, it will also provide the reasoning behind its conclusions and flag areas of uncertainty.

The personal physician AI will analyze long-term lifestyle data — including sleep, physical activity and behavioral patterns — gathered through wearables and home sensors to detect early signs of health deterioration.

The team plans to integrate all three AI systems into a unified medical AI platform covering the full cycle of care, from identifying disease causes through diagnosis and treatment to daily health management.

Validation for real-world clinical use will proceed in parallel. The team will collaborate with Samsung Medical Center, Vuno and Neurophet to conduct multi-institutional clinical trials and develop prototypes.

The project also focuses on nurturing early-career researchers in next-generation medical AI. Rather than treating them as mere participants, the program will have them lead independent projects as project leaders, with support including GPU infrastructure, clinical data access, hospital networks and opportunities for international collaborative research.

"We will bridge medical AI research with real clinical settings and everyday life, and create an environment where early-career researchers can take the lead and fully develop their research capabilities," Lee said.