Asan Medical Center (AMC) research team led by Professor In-Seob Lee of the Division of Stomach Surgery and Professor Kyung Won Kim of the Department of Radiology developed an AI algorithm predicting 5-year survival rate with an accuracy of approximately 80% for patients who underwent gastrectomy based on their 1-year postoperative results and health conditions.

The research team trained the AI with 65 large-scale variables from 3,220 patients who underwent gastric cancer surgery at AMC between 2003 and 2012. The variables included preoperative health information, surgery·chemotherapy·pathology information, and standard tests for recurrence follow-up observation results such as blood tests and computed tomography (CT).

After creating the AI algorithm, its internal validity was evaluated with data from 805 patients. As a result, it was found that the accuracy of 5-year survival rate prediction after gastric cancer surgery was approximately 76%. The AI algorithm was tested for external validity with data from 590 patients who underwent gastric cancer surgery at Ajou University Hospital between 2010 and 2012. The algorithm was found to predict the 5-year survival rate with an accuracy of 81%.

▲ Professor In-Seob Lee consulting a patient who underwent gastric cancer surgery

In addition, after an analysis of 65 types of patient data, it was found that changes in weight, muscle mass, fat mass, and nutritional status of the patients who underwent gastric cancer surgery were essential factors in predicting the 5-year survival. It was also demonstrated that deteriorating indicators such as weight and muscle mass loss and fat mass and NRI (nutritional risk index) increase were associated with a lower 5-year survival rate.

Professor In-Seob Lee said, “This research is significant for two reasons. First, we can now predict long-term outcomes after gastric cancer surgery through large-scale data analysis. Second, the research even demonstrated that the long-term survival of patients who underwent gastric cancer surgery is influenced by their self-correcting abilities, such as continuing muscle training and keeping high-protein diets.”

The research findings were recently published in the ‘Journal of Cachexia, Sarcopenia, and Muscle (IF=12.063),’ one of the prestigious international journals in Geriatrics. In addition, with the publication of this research article, Professor In-Seob Lee and Professor Kyung Won Kim were listed in Hanbitsa (People Glorifying Korea) of BRIC (Biological Research Information Center) as researchers who have published 3 or more articles in academic journals with an impact factor of 10 or higher in the last 3 years.