UC Davis Health’s AI model accurately predicts liver cancer risk, transforming patient care.

UC Davis Health’s team has pioneered an AI model to forecast liver cancer risks, focusing on hepatocellular carcinoma (HCC) in patients with metabolic dysfunction-associated steatotic liver disease (MASLD). Their study, shared in Gastro Hep Advances, showcases how this AI can guide doctors towards early, precise HCC risk assessments, promoting tailored patient care.

Understanding the stealthy MASLD—a fat accumulation condition in the liver linked to metabolic diseases—this research marks a significant step. By analyzing health data from over a thousand patients, the team employed machine-learning algorithms, particularly Gradient Boosted Trees, to identify HCC risks based on factors like liver fibrosis and other health indicators.

Their AI model achieved a 92.12% accuracy in predicting HCC occurrence, particularly benefiting patients traditionally deemed low-risk. This insight could lead to more effective screening and preventative strategies.

As they look to refine their model using natural language processing and Amazon’s Bedrock platform, the ultimate goal is to integrate this AI into healthcare systems, enhancing early detection and personalized care for liver disease patients.