Recorded 86% accuracy in predicting liver-related complications
The actual risk of occurrence is up to 30 times higher than that of the high-risk group and low-risk group.
AI has been developed to identify high-risk patients in advance who are more likely to develop liver-related complications after curing chronic hepatitis C.
The research team of Professors Kim Seung-up and Lee Hye-won of the Department of Gastroenterology at Severance Hospital announced on the 5th that they had developed AI that can select high-risk patients who may show liver-related complications after curing chronic hepatitis C.
Chronic hepatitis C is not transmitted through casual contact, but through contaminated blood such as needle sticks, acupuncture, tattoos, etc. It can be cured with most commonly consumed ‘Direct-acting Antiviral Agents (DAA)’.
But what happens after complete recovery is the problem. This is because some of those who have recovered may develop serious complications, such as hepatocellular carcinoma.
In cases where cirrhosis is not diagnosed through abdominal ultrasound, it is difficult to recognize the risk, so predicting it is very important.
First, AI was created using data from 913 hepatitis C patients without cirrhosis who were treated with antiviral treatment at Severance Hospital. AI predicts whether severe complications will occur by using the results of six blood tests that can check liver elasticity, age, gender, platelet count, and bilirubin level.
Next, the accuracy of predicting liver-related complications such as hepatocellular carcinoma, decompensated cirrhosis, liver transplantation, and death was tested using data from 1,264 patients in Hong Kong and France. The ‘C-index’ value, which means higher prediction performance the closer it is to 1, recorded 0.86, boasting an AI prediction accuracy of 86%. When only hepatocellular carcinoma was confirmed separately, the figure rose to 0.87.
In fact, the incidence of liver-related complications within 5 years in high-risk patients with AI-calculated values of 0.7 or higher was 3.2-6.2%, up to 30 times higher than the incidence rate (0.2-0.5%) in low-risk patients with AI-calculated values of less than 0.7.
Professor Kim Seung-up said, “Chronic hepatitis C can be easily treated with antiviral drugs, but even after cure, serious complications such as hepatocellular carcinoma can be experienced.” He added, “Using the AI developed in this study, high-risk groups that require regular follow-up even after cure are selected. “You can do it,” he said.
The results of this study were published in the Journal of Hepatology (IF 26.8), an academic journal of the European Association for the Study of the Liver.
Reporter Lee Jin-woo [email protected]
[ⓒ 세계일보 & Segye.com, 무단전재 및 재배포 금지]