More than a billion people worldwide carry excess fat in their livers, yet most will never know it until the damage is done. Now a wave of artificial intelligence tools is being developed to change that, scanning routine medical data to flag the disease early, when it can still be reversed. The stakes are enormous: fatty liver disease affects roughly 30% of adults globally, and three-quarters of people with cirrhosis are only diagnosed once the condition is life-threatening.
A Silent Epidemic Hiding in Plain Sight
Fat in a normal healthy liver is negligible. Disease sets in when fat exceeds 5% or 10% of the organ's weight. That accumulation triggers inflammation, cell damage, and the formation of scar tissue known as fibrosis. Left unchecked, progressive fat buildup can lead to liver failure, and it is linked to a higher risk of cardiovascular disease and various cancers. The condition typically develops without noticeable symptoms, which is why it is so rarely caught early.
Lifestyle changes can turn things around. Reducing alcohol, losing weight, exercising, and even drinking more coffee can reverse scarring and inflammation in the early stages. The problem is finding those early stages in the first place. Simple noninvasive tests exist, but they are rarely used even for high-risk groups such as people with obesity or type 2 diabetes. Physicians are already overwhelmed, and adding more testing to every visit is not sustainable.
Jeffrey Lazarus, a professor at the CUNY Graduate School of Public Health and Health Policy, sees a clear role for AI in closing that gap. "AI can retrospectively go through massive numbers of hospital visits and lab reports," he said. "You can use that to really prioritize who's at most risk."
From Chest X-Rays to Blood Tests: The New AI Arsenal
Researchers are attacking the problem from multiple angles. Last year, Osaka Metropolitan University published a study on an AI model that analyzes routine chest x-rays to detect fatty liver disease with 82% accuracy. That approach could turn an exam already performed for other reasons into a liver screening tool.
"The AI could pick up cases of excess liver fat, check for other risk factors such as if the person is overweight, has high cholesterol or type 2 diabetes, and then make a recommendation to the doctor," Lazarus said. "It could tell them, 'This might not have been what you were looking for, but this is what was picked up, and you should refer to hepatology or endocrinology who can do further tests.'"
Other tools work from blood work. The Danish health tech startup Evido developed LiverPRO, an AI algorithm that assesses liver fibrosis risk using age and nine routine blood-based biomarkers. Commercialized with Roche, LiverPRO outperformed the standard Fib-4 index in predicting serious liver problems in more than 470,000 middle-aged people.
Another model, ALADDIN, is a machine learning algorithm based on routine blood tests. Evaluated by an international collective of hepatologists earlier this year, it outperformed Fib-4 and other risk scores in identifying patients who could benefit most from resmetirom, a drug shown to be highly effective for moderate to advanced liver scarring. The GLP-1 medication semaglutide has also demonstrated strong results for the same stage of disease. AI-based tests could help select patients for these treatments without invasive biopsy.
The Limits of Current Testing
The Fib-4 index has been the workhorse of liver fibrosis screening for years. It computes a score from 0 to 6 based on age, two liver enzymes, and blood-clotting ability. A liver blood test is often part of the annual medical checkup in the US, so the data is frequently available. But Fib-4 has known weaknesses. It is less accurate in adolescents and seniors, and it raises concerns about false positives that can send worried patients for unnecessary follow-ups.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
A second-line test, the enhanced liver fibrosis test, measures two proteins involved in scar tissue and an enzyme that inhibits scar clearance. Using both Fib-4 and the enhanced liver fibrosis test improves diagnosis of advanced fibrosis by four-fold. That combination works, but it adds complexity to an already burdened system.
Jonathan Dranoff, professor of medicine at Yale University, believes AI can remove that burden. "You have to have something that can run in the background or it's easy to just hit a button and do it," he said. He foresees AI automating Fib-4 score calculations from routine blood tests, sparing physicians the manual work of pulling values and computing scores themselves.
Fixing the Bottlenecks in Primary Care
Paul Brennan, a specialty registrar in gastroenterology, hepatology, and internal medicine at the University of Dundee, sees AI tools like ALADDIN as a practical fix for a systemic problem. "These tools won't completely replace biopsies or imaging," he said. "But they could fix the bottlenecks in primary care where most fibrosis goes undetected. I expect them to be adopted as a smarter first pass, catching the moderate-risk patients that blunter tools may miss, and reducing unnecessary referrals to hepatologists."
That matters because the traditional focus has been on managing late-stage disease. "The liver is a very versatile and interesting organ because it can regenerate, fibrosis can be reversed, and you can be completely healthy again," Lazarus said. "But traditionally, we've focused more on late-stage care and trying to see how long we can keep a patient alive rather than finding them early and preventing the condition from advancing."
There is evidence that early detection changes behavior. Research in Denmark found that informing people they have liver fibrosis makes them more likely to commit to dietary and exercise regimes. "We're always looking to improve adherence to eating well and doing more physical activity," Lazarus said. "Telling someone that they might have or do have liver disease is one way to do that."
From Research to Practice
So far, AI in liver care has largely been confined to research. No major health system has deployed these tools at scale. But Lazarus is optimistic that the shift is coming. "I would love to say, let's go back through the electronic medical records across the various US health systems and find people before they have cirrhosis," he said.
The economic case is as strong as the medical one. Liver transplants are extraordinarily expensive in any country, especially the US. Catching the disease early, when lifestyle changes or medications can still work, would save health care systems vast sums of money. "Liver transplants are extraordinarily expensive in any country, but especially the US. So there's a lot of good humane and economic reasons to find people earlier on," Lazarus said.
The tools are not perfect. They will not replace the judgment of a hepatologist or the precision of an imaging study. But as a first pass, they could redirect the entire trajectory of care for a disease that has quietly become one of the world's most common health threats. With over a billion people already affected and numbers climbing, the question is no longer whether AI can help spot fatty liver disease. It is how quickly health systems will let it.