Hallym University Chuncheon Sacred Heart Hospital is embarking on the development of a microbiome-based AI solution that can diagnose chronic liver disease and predict its onset using a simple stool test.

On the 27th, the hospital announced that Professor Seok Ki-tae of its Digestive Disease Research Institute will lead a four-year project—from December through November 2025—to develop a precision medicine AI solution for the diagnosis and prediction of chronic liver disease, in collaboration with private companies including AIDOT, Sopung & Company, and E1 Diagnostics, among others.

Chronic liver disease, for which there are no appropriate therapeutics, is the fourth leading cause of death, following cancer, heart disease, and cerebrovascular disease. In Korea, the prevalence of liver disease is as high as 30%. While chronic liver disease can be treated appropriately if detected early, there is still no fast and accurate diagnostic method, and patients must undergo many steps, including liver function tests, ultrasound, and MRI.

To address this, the research team plans to combine microbiome research with next-generation sequencing (NGS) to identify alcoholic liver disease, and apply the findings to an AI algorithm.

It also plans to improve accuracy by selecting from 7,600 microbiome cases derived from human samples and training (train/validation) the model on 1,000 curated retrospective microbiome datasets.

The AI solution will diagnose by classifying the test subject’s liver status into four stages: normal (fatty liver), hepatitis, cirrhosis, and liver cancer.

In particular, through analysis of the examinee’s gut microbiome, it will be able to predict the likelihood that the liver will progress to cirrhosis, as well as the likelihood of a confirmed cirrhosis diagnosis.

Professor Seok Ki-tae of the hospital’s Division of Gastroenterology said, “Once this AI solution is completed, even primary and secondary medical institutions will be able to accurately diagnose or predict chronic liver disease through a simple stool test, contributing to reduced social costs arising from alcoholic liver disease.”

He added, “We will do our utmost to achieve innovative advances in prediction and early diagnosis of liver disease, including significantly reducing the time and cost required for microbiome and genomic analysis.”

Published: July 27, 2021 / New Daily Economy / Reporter Park Geun-bin