Multi-Region Radiomic Analysis Based on Multi-Sequence MRI Can Preoperatively Predict Microvascular Invasion in Hepatocellular Carcinoma

• DOI: 10.3389/fonc.2022.818681
• Publication Date: 2022-04-27T15:38:13Z
• Authors: Lanmei Gao; Meilian Xiong; Xiaojie Chen; Zewen Han; Chuan Yan; Rongping Ye; Lili Zhou; Yueming Li
• Keywords: nomogram; 03 medical and health sciences; machine learning; 0302 clinical medicine; Oncology; radiomics; microvascular invasion; magnetic resonance imaging; Neoplasms. Tumors. Oncology. Including cancer and carcinogens; hepatocellular carcinoma; RC254-282

Microvascular invasion (MVI) affects the postoperative prognosis in hepatocellular carcinoma (HCC) patients; however, there remains a lack of reliable and effective tools for preoperative prediction MVI. Radiomics has shown great potential providing valuable information tumor pathophysiology. We constructed validated radiomics models with without clinico-radiological factors to predict MVI.One hundred fifteen patients pathologically confirmed HCC (training set: n = 80; validation 35) who underwent MRI were retrospectively recruited. based on multi-sequence across various regions (including intratumoral and/or peritumoral areas) built using four classification algorithms. A model was individually combined generate fusion by multivariable logistic regression.Among models, T2WI arterial phase (T2WI-AP model) volume liver-HCC interface (VOIinterface) exhibited best predictive power, AUCs 0.866 training group 0.855 group. The good efficacy (AUC: 0.819 0.717, respectively). showed excellent ability 0.915 0.868, respectively), outperforming both T2WI-AP sets.The multi-region achieves an enhanced individualized risk estimation MVI patients. This may be beneficial tool clinicians improve decision-making personalized medicine.