Hallym University Successfully Concludes 'Medical Data Utilization Competition'... AI Model for Predicting Treatment Outcomes in Elderly Patients Wins Grand Prize

Hallym University SEPO Team Predicts Treatment Outcomes for Elderly Hypertensive Patients
Grand Prize at Medical Data Utilization Competition with AI Model
Building an AI Healthcare Education and Research Ecosystem through University-Medical Institution Collaboration

Hallym University (President Choi Yanghee), the Medical AI Convergence Talent Development Project Group, the Software-Centered University at Hallym University, and Hallym University Medical Center (Director Kim Yongseon) held the awards ceremony for the '2nd Medical Data Utilization Competition (H-iDEA, Hallym innovation BigData Exploration & Analysis) 2025' on November 12 at the Ilsong Culture Hall, Annex 2, Hallym University Sacred Heart Hospital.

Commemorative photo of the winners of the 2nd Medical Data Utilization Competition 'H-iDEA' in 2025. Provided by Hallym University

This competition was jointly organized by Hallym University Medical Center, the Medical AI Convergence Talent Development Project Group at Hallym University, and the Software-Centered University Project Group at Hallym University, targeting Hallym University students to promote the development of innovative technologies based on medical data and to discover and nurture outstanding talent in the field of medical artificial intelligence (AI).

A total of 24 teams, comprising 47 students, participated in the event, demonstrating high interest in the field of medical AI from students across diverse disciplines such as artificial intelligence, medicine, natural sciences, and social sciences. The participating teams utilized anonymized clinical data provided through 'HERO,' Hallym University Medical Center's next-generation data lake cloud platform, to demonstrate various possibilities in medical AI. The winning teams were selected through evaluation of their final reports by experts in medicine, data, and information science, resulting in one grand prize, two excellence awards, and four merit awards.

The grand prize was awarded to the SEPO team (Choi Seungah, Oh Wonjun, Han Subin from Hallym University College of Medicine) for their project, 'Development of a Predictive Model for Treatment Failure Risk in Discharged Elderly Hypertensive Patients and Study on Its Application to the Medical Social Work Team.'

This team designed a predictive model to identify high-risk elderly patients with a high likelihood of treatment failure at an early stage, aiming to improve treatment adherence and reduce readmission rates among elderly hypertensive patients. Through this, the medical social work team can select elderly patients in need of intensive care interventions and utilize limited resources more efficiently.

The SEPO team applied SHAP (SHapley Additive exPlanations) analysis to multidimensional data of admitted and discharged patients, confirming the contribution of each variable to prognosis prediction. As a result, they identified meaningful variables that can explain functional recovery and post-discharge status of patients, such as admission date, gender, family relationships, discharge type, bathing method, and emotional state.

Choi Seungah, leader of the SEPO team, stated, "We aimed to analyze and build a model focusing on the impact of socioeconomic factors on disease prognosis in elderly patients," and added, "I will continue to learn, never forgetting the essence of realizing better healthcare for patients."

Suh Younggyun, Director of the Big Data Center (Professor of Family Medicine at Hallym University Sacred Heart Hospital), remarked, "Utilizing medical data is challenging, but when done well, its significance is beyond imagination. We will strive to apply the excellent ideas discovered in this competition to real medical settings and create a joint research ecosystem so that medical data can be fully utilized through various competitions and collaborations using HERO."

Lee Jaejun, Head of the Medical AI Convergence Talent Development Project Group (Director of Hallym University Chuncheon Sacred Heart Hospital), commented, "The creative attempts by students based on medical data are very encouraging. We will continue to expand practical education and research opportunities in collaboration with medical institutions to nurture world-class medical AI convergence talent."

Meanwhile, Hallym University Medical Center established the Big Data Center under the Digital Medical Innovation Research Institute (DIDIM) in 2021 and is carrying out major national projects such as the 'Medical Data-Centered Hospital Support Project' and the 'K-CURE Clinical Data Network Construction Project.' In addition, by developing its own data lake cloud platform, 'HERO,' and becoming the first medical institution to simultaneously obtain certification for 'medical data content and management system,' it is leading data-driven medical innovation.

The Medical AI Convergence Talent Development Project Group at Hallym University, supported by the Korea Health Industry Development Institute under the Ministry of Health and Welfare, operates practical education programs aimed at nurturing next-generation medical AI talent based on the convergence of medicine and AI technology. The group plans to take the lead in building an education and research ecosystem in the AI healthcare field through close cooperation with universities, medical institutions, and industry.

This content was produced with the assistance of AI translation services.