Abstract
Hepatocellular carcinoma (HCC) is a leading cause of cancer mortality, with advanced disease resistant to conventional therapies. Targeted combined immunotherapy, such as ICIs with TKIs or anti-VEGF agents, has improved outcomes, but heterogeneous responses necessitate predictive biomarkers. Molecular imaging (MRI, PET/CT, CT, CEUS) provides non-invasive tumor immune microenvironment (TIME) insights for stratification and monitoring. This review synthesizes advances, evidencing AUCs > 0.85 in response prediction, while critiquing limitations like artifacts and proposing AI-multimodal fusion. We outline a roadmap for precision oncology, enhancing efficacy and personalization in HCC.