AI Blood Test Uncovers Hidden Liver Disease Before Symptoms
Johns Hopkins researchers have developed an AI-powered liquid biopsy that analyzes genome-wide patterns of cell-free DNA fragments to detect liver fibrosis and cirrhosis years before symptoms, using about 40 million fragments across thousands of genomic locations including repetitive DNA regions. The fragmentome-based approach, aided by machine learning, aims to identify early liver disease and potentially other chronic conditions beyond cancer. The liver-disease classifier is still a prototype and not yet clinically available; validation is ongoing. If validated, earlier detection could enable interventions to reverse fibrosis and reduce liver-cancer risk for millions at elevated risk in the U.S.
