Artificial intelligence in the prediction of progression and outcomes in viral hepatitis

Because of improvements in image digitization and the popularization of electronic health big data, research involving artificial intelligence (AI) in viral hepatitis has grown rapidly. AI can be used to diagnose and predict the disease progression of viral hepatitis, such as liver fibrosis, cirrhosis, and hepatocellular carcinoma, and identify insignificant microscopic diseases automatically and predict prognoses using multiomics, histopathology, and radiology data. AI is expected to affect the diagnosis, prediction, and precision medicine of hepatitis significantly, helping physicians manage patients and reduce medical error. However, there are still hurdles to overcome in applying AI to viral hepatitis. First, AI requires a large number of high-quality training datasets and considerable computing power. In addition, clinical application of the AI-generated black box model remains to be resolved. Furthermore, a suitable platform needs to be established to conduct simultaneous external validation studies of multiple AI models to compare their performance. Finally, AI needs to be used to assist rather than replace doctors. Appropriate ways should be found for experts to collaborate with AI.