Radiomic Analysis in Hepatocellular Carcinoma: Prospects and Clinical Benefits (Review)

Abstract

Introduction: Hepatocellular carcinoma (HCC) is associated with high mortality due to challenges in early diagnosis, the subjectivity of imaging assessments, and the lack of reliable, non-invasive methods for predicting tumor aggressiveness and treatment response.
Purpose: To evaluate the potential and evidence base of radiomic image analysis in addressing key clinical challenges in HCC.
Materials and methods: A review of current scientific literature, including Russian and international studies from 2023 to 2025, was conducted. The methodology of radiomics and its effectiveness were analyzed for the following applications: differential diagnosis of HCC, prediction of microvascular invasion (MVI), and evaluation or prediction of the effectiveness of transarterial chemoembolization (TACE) and radiofrequency ablation (RFA).
Results: Radiomic models demonstrated high accuracy (with sensitivity up to 96 %) in the differential diagnosis of HCC. The integration of 3D radiomic features with clinical and laboratory data enabled the prediction of MVI, achieving sensitivity rates of 76–82 % and specificity of 82–85 %. Combined clinical–radiomic models showed strong performance in predicting response to TACE (AUC up to 0.92) and RFA (AUC up to 0.87), as well as in assessing recurrence risk—outperforming traditional approaches.
Conclusion: Radiomic analysis is a promising non-invasive tool for assessing HCC aggressiveness and guiding treatment selection. It outperforms conventional imaging in predicting MVI and therapeutic response to local treatments such as TACE and RFA.