Development of Gachon University Gil Hospital 'AI gastric cancer prediction software obtains certification from the Ministry of Food and Drug Safety

Aug 28, 2024

The artificial intelligence gastric cancer prediction software being developed by Gachon University Gil Hospital (Hospital Director Kim Woo-kyung) recently obtained permission to manufacture medical devices from the Ministry of Food and Drug Safety.

Gachon University Gil Hospital is participating in the 'Doctor Answer 2.0' for gastric cancer prediction and early detection as part of the 'AI Precision Medical Solution' development project organized by the Ministry of Science and ICT and the Information and Communication Industry Promotion Agency. Following the colorectal cancer prediction and diagnosis software (Dr. Answer 1.0) developed in 2021, stomach cancer has been expanded to be applied.

Among the Doctor Answer 2.0 projects, gastric cancer diseases are being developed by dividing software that predicts gastric cancer using artificial intelligence and early diagnosis of gastric cancer during gastric cancer endoscopy. Among them, AI-Doctor for Gastric Cancer (AI-Doctor for Gastric Cancer) has recently been certified by the Ministry of Food and Drug Safety to predict the possibility of gastric cancer through medical records (EMR) and interviews of patients. Gachon University Gil Hospital jointly developed this software by signing a partnership with PCT, a company specializing in information and communication technology.

The software uses the subject's blood test results, health checkups such as Helicobacter tests, electronic medical record data, and survey results on lifestyle to derive the risk rate of gastric cancer.

More than 30,000 EMR big data including lifestyle, health checkups, endoscopy and biopsy are secured and multi-faceted analysis of gastric cancer is conducted.

"Gastric cancer can be diagnosed early through endoscopy, but sometimes it does not comply with the test due to patients' anxiety factors such as old age and constitution," said Kim Kyung-oh, a professor of gastroenterology who is in charge of the study. "In this case, we aim to calculate the quantified risk rate of gastric cancer to increase the patient's endoscopic compliance and ultimately to diagnose gastric cancer early."

Risk groups who do not undergo examination for reasons such as 'must be tested' or 'endoscopy is burdensome' can analyze the risk rate and provide it as a visual result only with their electronic medical records and questionnaires (questionnaires).

Gastric cancer is the number one cancer incidence in Korea, beating lung cancer, liver cancer, and colon cancer, and the number of gastric cancer patients treated over the past five years has increased by 6.9%. In Korea, about 30,000 people occur every year, and 70% of them are detected early. The most common routes are found through health checkups such as national cancer checkups, and as of 2021, the screening rate for gastric cancer checkups was 62.6%, which is lower than the highest liver cancer rate of 72.9%.

Kim Woo-kyung, head of Gachon University Gil Hospital, said "The goal of the Doctor Answer project is to use artificial intelligence technology to alert patients to health care and detect cancer early. Doctor Answer SW, which can diagnose gastric cancer early and improve treatment results, can be widely used in the clinical field."



Development of Gachon University Gil Hospital 'AI gastric cancer prediction software obtains certification from the Ministry of Food and Drug Safety


bellho@sportschosun.com