Development of a Medical-Specialized Large-Scale AI Foundation Model Based on Multimodal Data
Securing Global Compatibility Through Application of International Medical Data Standards
"We Will Prove the Technological Capabilities of Korean Medical AI"
Deepnoid, a leading first-generation medical AI company in Korea, announced on July 31 that it has been selected as the lead research and development institution for a large-scale AI technology development project worth a total of 11.6 billion KRW, supervised by the Ministry of Trade, Industry and Energy.
Deepnoid was finally selected for the "2025 First Biohealth R&D Project in the Digital Healthcare Sector," which is supervised by the Ministry of Trade, Industry and Energy and the Korea Evaluation Institute of Industrial Technology (KEIT). The agreement process has also been completed. The selected project is "Development of Medical-Specialized Multimodal Large-Scale Generative AI Technology." Under the overall supervision of the Korea Testing Laboratory (KTL), Deepnoid will serve as the lead research and development institution. WinU, the Catholic University of Korea Industry-Academic Cooperation Foundation, the Yonsei University Industry-Academic Cooperation Foundation, and the National Cancer Center will participate as joint research institutions.
This year, Deepnoid accumulated experience in generative AI technology development, including the creation of "M4CXR," a generative AI-based model for drafting chest X-ray (CXR) reports. Through this project, Deepnoid, as the lead research and development institution, will directly plan and execute the entire process of developing a large-scale AI foundation model.
Starting this year and continuing for the next five years, the government will provide approximately 8.55 billion KRW in support. Deepnoid will receive 4.98 billion KRW in government funding to conduct research and development on a medical-specialized large-scale AI foundation model optimized for the domestic clinical environment, based on multimodal data such as medical imaging, biosignals, and EMR.
Deepnoid aims to go beyond simple model development to implement medical imaging report generation technology, and to develop and commercialize technology that can be used in the global market. The company plans to ensure global compatibility by applying international medical data standards such as HL7, FHIR, and SNOMED CT from the development stage.
Ultimately, the project aims to: ▲ improve the efficiency of medical professionals ▲ enhance diagnostic accuracy ▲ reduce social medical costs ▲ promote the digital transformation of Korea's medical infrastructure ▲ and secure global competitiveness for the domestic medical industry.
Choi Woosik, CEO of Deepnoid, said, "Through this project, we have expanded our medical AI capabilities to large-scale generative models," adding, "As the competition in the medical AI market is fierce, the ability to demonstrate commercialization is crucial." He continued, "We will successfully carry out this project and prove the technological capabilities and competitiveness of Korean medical AI."
Deepnoid is currently proceeding with clinical trial procedures to obtain approval from the Ministry of Food and Drug Safety for "M4CXR." The company is also developing and operating various medical AI solutions, such as "Deep:Neuro," an AI solution for assisting in the interpretation and diagnosis of cerebral aneurysm images. With this project as a momentum, Deepnoid is pursuing both technology advancement and market expansion.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
