Sang Min Park, CEO of XAIMED
Its flagship product is the ‘Fundus-Based Arteriosclerosis AI Solution’, a medical AI solution for identifying opportunities
Development of technology capable of predicting both carotid and coronary atherosclerosis using only fundus photography
Its flagship product is the ‘Fundus-Based Arteriosclerosis AI Solution’, a medical AI solution for identifying opportunistic. Fundus photography is a basic test for diagnosing ophthalmic diseases such as diabetic retinopathy, glaucoma, and macular degeneration. XAIMED has developed technology that identifies hidden vascular health information within fundus images. This enables the simultaneous prediction of carotid and coronary atherosclerosis using only noninvasive fundus photography.
The ‘Fundus-Based Arteriosclerosis AI Solution’ has demonstrated predictive power that is independent of the Framingham Risk Score, which is an existing indicator of cardiovascular risk. Furthermore, it has proven its potential to play a crucial role in prompting early patient management and lifestyle improvements, as patients categorized as high risk exhibited significantly higher actual cardiovascular disease mortality rates.
Screening institutions can obtain images using only basic fundus photography instead of expensive imaging tests. This reduces the burden on patients and healthcare professionals. Furthermore, providing guidance for high-risk groups to undergo detailed examinations based on screening results enhances diagnostic efficiency and focused judgments. The ‘Fundus-Based Arteriosclerosis AI Solution’ is an innovative solution that enables the early assessment of risk for major cardiovascular diseases, such as stroke and myocardial infarction.
XAIMED is also developing an ‘AI for osteoporosis prediction based on chest X-ray imaging.’ Using Seoul National University Hospital’s extensive clinical database, research has proven that a single chest X-ray can predict bone density in the lumbar spine and femur area, enabling early osteoporosis diagnosis. Additional verification studies are currently underway with other medical institutions. XAIMED is expanding its opportunistic diagnosis AI pipeline by utilizing diverse medical modalities, including fundus images, chest X-rays, abdominal CT scans, electrocardiograms (ECGs), skin conductance, and PPG sensors. Furthermore, XAIMED is developing time-synchronized bio-integrated sensor equipment to prepare a next-generation opportunistic diagnosis AI solution based on multiple biosignals.
“XAIMED’s world-class explainable AI (XAI) technology provides a strong competitive advantage in developing the optimal solution for diagnosing opportunities. To overcome the limitation whereby even highly accurate AI is often not understood, XAIMED has developed methodologies that maximize both diagnostic performance and explainability. Through the ‘Adversarial Counterfactual Explanation’ methodology, we help physicians to intuitively understand the AI interpretation process. By developing methodologies to quantify scientific validity and explainability, we have created the ‘Auditing AI’ platform—a medical AI development process that maximizes both diagnostic performance and explainability. This approach significantly reduces the time and cost required to develop ‘optimal medical AI’ by evaluating not only prediction accuracy but also the transparency with which the AI’s decision-making process can be explained.”
The company is rapidly and efficiently developing medical AI solutions that can diagnose arteriosclerosis, cardiovascular disease, osteoporosis, and other conditions. These solutions use common examination data, such as fundus images, chest X-rays, abdominal CT scans, and ECGs, based on this platform. It works with big screening centers from the beginning to test the solutions, making sure they can be used right away in hospitals. This results in a high adoption rate among medical institutions.
XAIMED’s ‘Fundus-Based Arteriosclerosis AI Solution’ was developed using data from over 30,000 patients at Seoul National University Hospital. The Ministry of Food and Drug Safety (KFDA) granted authorization and permission in 2024 after verification testing on 100,000 examinees from the Korea Association of Health Promotion (KAHP), which conducts annual health examinations for approximately 2.5 million people. The solution is currently formally implemented across all 17 KAHP branches, generating revenue and proving its clinical utility and reliability. Collaborations with multiple medical device distributors enable its use in various medical institutions. Starting in September 2025, the solution was selected for the New Medical Technology Evaluation Deferral Program by the National Evidence-based Healthcare Collaborating Agency (NECA), which is under the Ministry of Health and Welfare. This selection enables the solution’s legal clinical use as a non-benefit item in medical institutions nationwide.
Regarding future plans, Park stated, “Based on explainable AI (XAI) technology, XAIMED will grow into a world-class innovative company that develops an optimal opportunistic diagnosis AI solution satisfying both diagnostic performance and scientific validity, while finding new digital biomarkers.”
Founded: April 2020
Main business: development and supply of explainable AI solutions and bio-integrated sensor intelligent solutions
Achievements: Fundus-based cardiovascular AI solution obtained KFDA permission for medical devices and selected for the New Medical Technology Evaluation Deferral Program; Formally adopted across all 17 KAHP branches nationwide, commercialized and generated from 2025; Obtained GMP certification; Obtained permission for manufacturing medical devices; Selected for multiple government R&D projects (including CES 2024 participation); Won the Korean Academy of Medical Sciences’ Wunsch Medical Award for Young Medical Researchers, the Pfizer Medical Award from the National Academy of Medicine of Republic of Korea, and the Bumsuk Award from the Bumsuk Academic Scholarship Foundation, etc.
reporter jinho lee
jinho2323@hankyung.com
© 한경매거진&북, 무단전재 및 재배포 금지