"Medical AI to Lead Personalized Health Management... Should Be Seen as a 'Signal,' Not a 'Diagnosis'"

As artificial intelligence (AI) medical technology is expected to rapidly shift health checkups from "disease detection" to "personalized health management," experts have pointed out the need to improve the accuracy and validity of screenings in real clinical settings and to address sensitive issues such as the security of personal information.

On the 15th, at the Korea Press Center in Jung-gu, Seoul, An Jihyun, Senior Executive Researcher at the Korea Medical Institute (KMI), gave a presentation titled "How AI-Implemented Health Screening Centers Have Changed" at the symposium "A Paradigm Shift in Medical Care: The Present and Future of AI Health Screening." Photo by Jo Inkyung

원본보기 아이콘

During the symposium "A Paradigm Shift in Medical Care: The Present and Future of AI Health Screening," hosted by the Korea Medical & Bio Journalists Association on April 15, experts discussed the changes brought by AI in the field of health screening and explored ways to advance a Korean-style AI screening model.

Kang Daehee, Professor in the Department of Preventive Medicine at Seoul National University College of Medicine, who delivered the keynote lecture, identified "4P Medicine" (Predictive, Preventive, Personalized, and Participatory) as the core value of future healthcare and proposed ways to utilize AI-based health screenings to realize this goal. According to Professor Kang, AI can be used to optimize test items by selecting candidates and identifying high-risk groups before screenings; during screenings, it can standardize image and pathology interpretations and reduce differences among examiners; and after screenings, it can provide specific guidelines and personalized follow-up care.

Professor Kang stated, "AI has already been widely adopted throughout the medical field," and predicted, "In the future, AI-based health screening will shift from one-time checkups to a continuous, subscription-based health management system."

Concrete achievements from actual clinical settings were also shared. An Jihyun, Senior Executive Researcher at the Korea Medical Institute (KMI), explained, "By using AI-assisted systems for colonoscopy, we reduce missed lesions and, for example, can predict cardiovascular disease risk using only fundus photographs, thereby creating new medical value by integrating AI into existing tests." An further highlighted, "In particular, predicting cardiovascular disease through fundus photography is significant in maximizing the effectiveness of 'opportunistic screening' without additional radiation exposure," and evaluated, "Collaboration between AI and medical professionals can prevent burnout among healthcare workers and create a mutually beneficial model for precision medicine."

Jung Myunghoon, Country Manager of Guardant Health Korea, introduced "liquid biopsy," a next-generation screening technology, as a promising method for early cancer detection. Jung said, "Through epigenomics analysis, we are able to detect even trace amounts of genetic mutations in the blood," and claimed, "With just a single blood draw, multiple cancers can be identified simultaneously and the site of occurrence can even be predicted through multi-cancer detection (MCD), thereby addressing existing blind spots in screening."

However, there were also critical views cautioning against an overreliance on AI screening. Kim Hyungjin, Professor at Samsung Medical Center's International Health Care Center, warned, "AI results are more like 'signals' that require further confirmation, not definitive diagnoses, so if they are misinterpreted as diagnoses, this could lead to unnecessary tests and invasive procedures." He also pointed out that AI reflects the biases of its training data and that large-scale datasets do not necessarily guarantee fairness. Professor Kim emphasized, "While there is no need to distrust AI health screenings, it is essential to interpret the results critically and ask questions to ensure safe medical care."

On the 15th, at the Korea Press Center in Jung-gu, Seoul, during the symposium titled 'A Paradigm Shift in Healthcare: The Present and Future of AI Health Screening,' panelists including Minwoo Cho, Professor of Preventive Medicine at Ulsan University College of Medicine, Hyeun Jeong, Director of Health Promotion Policy at the Ministry of Health and Welfare, Myunghee Park, CEO of Consumer Together, and Suhyun Lee, CEO of Tessar, are participating in a panel discussion. Photo by Inkyung Jo

원본보기 아이콘

Myunghee Park, CEO of Consumers Together (and Honorary Professor at Dongguk University), expressed concerns from a medical consumer's perspective: "AI health screening is like having a smart primary care physician by your side, but at the same time, psychological side effects can arise, such as excessive stress over the risk figures for future disease occurrence predicted by AI and concerns about data safety." Park also pointed out, "There are numerous unresolved issues, including the leakage of sensitive information such as genomic data and disease history, the possibility that collected health data could be misused for insurance denial or premium increases, and the uncertainty about who is responsible in the event of an AI misdiagnosis." She added, "At this point, AI screenings should be used as an aid to doctors' decision-making, and a thorough security system should be prioritized."

Suhyun Lee, CEO of Tessar, commented, "In comprehensive health screening settings, where it is virtually impossible for medical professionals to meticulously interpret the results of numerous tests and write reports every day, smart screening AX (AI transformation) solutions can enable early disease detection, address the limitations of follow-up care due to staffing shortages, and allow for detailed interpretation and systematic health management that would be difficult to achieve in large-scale screenings." Lee further advised, "It is important to ensure that AI adoption is natural for frontline staff by building on security infrastructure, maximizing AI precision, and providing function modules optimized for hospital workflows."

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.