AI Could Help Detect Fatty Liver Disease Early, Experts Say
Evido
Osaka Metropolitan University
Fatty liver disease affects about 30% of adults worldwide but often goes undetected until late stages. AI tools analyzing electronic health records, routine blood tests, and x-ray images could help identify at-risk patients early, potentially reversing damage and saving costs. Researchers and startups like Evido are developing AI models that outperform traditional risk scores.
Fatty liver disease, affecting roughly 30% of adults globally, often progresses silently, with three-quarters of cirrhosis cases diagnosed only when life-threatening. Researchers like Jeffrey Lazarus of CUNY propose using AI to comb through electronic health records and prioritize at-risk patients. Simple noninvasive tests like Fib-4 exist but are underused; AI could automate their calculation from routine blood data. A 2024 Osaka Metropolitan University study showed an AI model could identify fatty liver disease from chest x-rays with 82% accuracy. Danish startup Evido's algorithm LiverPRO, commercialized with Roche, outperforms Fib-4 in predicting liver problems. Another AI model, ALADDIN, helps select patients for resmetirom treatment. These tools aim to fix diagnostic bottlenecks in primary care, reducing unnecessary referrals and enabling early intervention that can reverse liver damage.
- Abbreviations
- GLP-1 = Glucagon-Like Peptide-1 — глюкагоноподобный пептид-1
- Fib-4 = Fibrosis-4 index — индекс фиброза-4
Source: Wired AI —
original