AI Blood Test Spots Silent Liver Disease Years Early

A simple blood test could soon catch deadly liver disease before it causes any symptoms, giving doctors years to intervene and save lives.

Researchers at Johns Hopkins Kimmel Cancer Center developed an AI-powered test that analyzes tiny DNA fragments floating in your bloodstream. The system detected early liver fibrosis and cirrhosis with remarkable accuracy by examining how DNA pieces break apart across the entire genome.

The breakthrough uses "fragmentome technology" to read patterns in about 40 million DNA fragments per test. Unlike traditional liquid biopsies that search for cancer mutations, this approach captures a wider picture of your body's health by looking at how cells package and release DNA.

"For many of these illnesses, early detection could make a profound difference," says Dr. Victor Velculescu, who co-led the study published in Science Translational Medicine. "Liver fibrosis is reversible in its early stages, but if left undetected, it can progress to cirrhosis and ultimately increase the risk of liver cancer."

The timing matters enormously. Current blood tests for liver fibrosis lack sensitivity and catch cirrhosis only about half the time. Many people don't realize they have liver disease until serious damage has already occurred.

The research team studied DNA samples from 1,576 people with liver disease and other conditions. Machine learning algorithms crunched the massive dataset to identify fragmentation patterns linked to disease stages.

Why This Inspires

Roughly 100 million Americans live with liver conditions that increase their risk of cirrhosis and cancer. Most have no idea they're at risk until it's too late to reverse the damage.

This test could change that reality. By catching disease in its earliest stages when it's still reversible, doctors could treat underlying problems before they become life-threatening.

The technology may extend far beyond liver disease too. The researchers discovered their approach works for detecting multiple chronic conditions, not just cancer. That means one blood test could potentially screen for various diseases simultaneously.

The team stumbled onto this discovery while studying liver cancer patients in 2023. They noticed subtle DNA signals in people with early fibrosis, even when other markers looked normal. That observation sparked an entirely new research direction.

Machine learning made the breakthrough possible by processing enormous amounts of data that would overwhelm traditional analysis methods. The AI identified patterns invisible to human researchers.

Early intervention could prevent thousands of liver cancer cases each year. More importantly, it could give people decades of healthy life they'd otherwise lose to undetected disease.