Artificial Intelligence in Non-Alcoholic Fatty Liver Disease and Fibrosis: A Narrative Review.

Abstract

Non-Alcoholic Fatty Liver Disease (NAFLD) is a prevalent chronic liver condition that can progress to non-alcoholic steatohepatitis, fibrosis, cirrhosis, and hepatocellular carcinoma. While liver biopsy remains the gold standard for diagnosis, its invasiveness and cost limit its routine use. Recent advances in Artificial Intelligence (AI), particularly machine learning and deep learning, have created opportunities for accurate, non-invasive, and scalable assessment of NAFLD and related fibrosis. This narrative review summarizes recent studies applying image-based AI techniques, including convolutional and recurrent neural networks, as well as multimodal models combining imaging and clinical data. These approaches enhance the detection and grading of hepatic steatosis and fibrosis, improve diagnostic accuracy compared with conventional imaging or scoring systems, and enable standardized, cost-effective workflows using widely available modalities such as ultrasound and magnetic resonance imaging. Challenges remain, including the need for large, well-annotated datasets, interpretability of deep learning models, and mitigation of algorithmic bias. Despite these limitations, AI-assisted imaging holds substantial promise for earlier diagnosis, risk stratification, and personalized patient monitoring for NAFLD. Successful translation into clinical practice will require multidisciplinary collaboration, robust validation across diverse populations, and careful attention to ethical considerations such as data privacy and fairness that ultimately support improved patient outcomes and more efficient management of liver disease.