MRI-based radiomics models for prediction of vessels that encapsulate tumor clusters and microvascular invasion in hepatocellular carcinoma

Motivation: Microvascular invasion (MVI) and vessels that encapsulate tumor clusters (VETC) are significantly associated with poor prognosis. Sorafenib could improve the prognosis of hepatocellular carcinoma (HCC) patients with either MVI or VETC. Preoperative identification of VM-HCC (VETC+/MVI+, VETC+/MVI-, VETC-/MVI+) is essential for treatment decision-making. Goal(s): To develop multiregional MRI-based radiomics models and a clinical-radiomics nomogram for predicting VM-HCC, and to explore the prognostic value of these radiomics models. Approach: Retrospective study. Results: The peritumor radiomics model and the radiomics-clinical nomogram could effectively identify VM-HCC preoperatively. VM-HCC predicted by the peritumor model was an independent predictor of recurrence-free survival. Impact: Preoperative non-invasive identification of VM-HCC assists clinicians in developing personalized treatment strategies, thereby improving patient prognosis.