Application of Imaging in Predicting Treatment Efficacy of Transarterial Interventional Therapy for Hepatocellular Carcinoma: From Conventional Features to Multimodal Integrated Models

Abstract

Hepatocellular carcinoma (HCC), a major global cause of cancer death, faces treatment hurdles due to tumor heterogeneity and
therapy resistance. While transarterial therapy is key for unresectable HCC, patient outcomes vary widely, underscoring the need for reliable imaging biomarkers to predict treatment response and survival. This review systematically analyzes predictive biomarkersfrom
conventional radiological criteria and functional imaging to AI-enhanced radiomics-deep learning fusion modelshighlighting their clinical value and identifying research gaps. Multimodal integration emerges as the most promising approach for precise early-response prediction and long-term survival stratification, though current limitations include inconsistent imaging standards and unverified short-to-longterm prediction links.