Pusan National University has been selected for the Korean government’s AI Star Fellowship Program, led by the Ministry of Science and ICT (MSIT) and the Institute for Information & Communication Technology Planning & Evaluation (IITP). The program is designed to support top-tier early-career AI researchers and foster innovative research capable of driving future advances in artificial intelligence.

A total of KRW 12.1 billion will be invested in the project over approximately six years, from the second half of 2026 through December 2031. The funding includes KRW 11 billion from the national government, KRW 550 million from the Busan Metropolitan Government, and KRW 550 million in matching funds from PNU.

Led by Professor Giltae Song of the School of Computer Science and Engineering, the project will bring together PNU faculty specializing in AI from the School of Computer Science and Engineering, School of Electrical and Electronics Engineering, and School of BioMedical Convergence Engineering to develop a next-generation multimodal medical foundation model and a comprehensive precision medicine platform with broad applicability.

Key research areas include:

- AI-powered auftonomous robotic ultrasound scanning and 3D diagnostic and surgical-assistance technologies

- high-reliability diagnostic and reasoning technologies based on multimodal LLMs integrating medical imaging, language, and domain knowledge

- disease biomarker extraction using nanosensor data and pathology imaging

The technologies developed through the project are expected to undergo real-world validation and application at local medical institutions, helping move advanced medical AI research closer to clinical practice. PNU will also collaborate with major AI and healthcare organizations, including Samsung Medison, 42Maru, and Seegene Medical Foundation, to accelerate the commercialization of research outcomes and strengthen the training of industry-ready AI talent.

Through the AI Star Fellowship Program, PNU aims to cultivate world-class early-career AI researchers, build a stronger medical AI ecosystem in southeastern Korea, and accelerate AI Transformation (AX) across local industries.

What is a multimodal medical foundation model?

A multimodal medical foundation model is a large-scale AI model designed to learn from and integrate multiple types of healthcare data, such as medical images, clinical text, electronic health records, genomic information, and laboratory data.

Photo: Overview of the AI Star Fellowship project.