Hepatocellular carcinoma (HCC) is one of the most common malignant tumors worldwide. Microvascular invasion (MVI) is considered the major risk factor for postoperative recurrence and metastasis in HCC. The diagnosis of MVI relies on the postoperative pathological assessment of the tumor tissues. Seeking non-invasive methods and biomarkers for evaluation of MVI before surgery has important clinical implications for guiding surgical treatment and improving patients' survival. Recent studies have reported the applications of radiomics technique in prediction of MVI in HCC and showed promising results. Herein we summarized the research progress in CT- or MRI-based radiomics models for prediction of MVI in HCC to provide helpful thinking for further research in this field.肝细胞癌是全球最常见的恶性肿瘤之一,微血管侵犯(MVI)是术后复发及转移的重要原因。MVI的诊断依靠术后肿瘤的病理学检查,术前寻求无创评估MVI的方法及生物标志物,对指导手术治疗和提高患者生存具有重要临床意义。近年来不少研究报道了利用影像组学预测肝细胞癌MVI,结果表明该技术在本领域具有一定的应用前景。现总结了近年来基于CT或MRI图像影像组学模型预测MVI研究结果,以期为该领域的进一步研究提供有益的思考。.