Cen Yong-Yi, Nong Hai-Yang, Huang Xiao-Xiao, Lu Xiu-Xian, Pu Chang-Hong, Huang Li-Hong, Zheng Xiao-Jun, Pan Zhao-Lin, Huang Yin, Ding Ke, Huang De-You
Guangxi Clinical Medical Research Center for Hepatobiliary Diseases, Affiliated Hospital of Youjiang Medical University for Nationalities, Baise 533000, Guangxi Zhuang Autonomous Region, China.
Department of Radiology, Affiliated Hospital of Youjiang Medical University for Nationalities, Baise 533000, Guangxi Zhuang Autonomous Region, China.
World J Gastroenterol. 2025 Aug 14;31(30):109186. doi: 10.3748/wjg.v31.i30.109186.
Microvascular invasion (MVI) is an important prognostic factor in hepatocellular carcinoma (HCC), but its preoperative prediction remains challenging.
To develop and validate a 2.5-dimensional (2.5D) deep learning-based multi-instance learning (MIL) model (MIL signature) for predicting MVI in HCC, evaluate and compare its performance against the radiomics signature and clinical signature, and assess its prognostic predictive value in both surgical resection and transcatheter arterial chemoembolization (TACE) cohorts.
A retrospective cohort consisting of 192 patients with pathologically confirmed HCC was included, of whom 68 were MVI-positive and 124 were MVI-negative. The patients were randomly assigned to a training set (134 patients) and a validation set (58 patients) in a 7:3 ratio. An additional 45 HCC patients undergoing TACE treatment were included in the TACE validation cohort. A modeling strategy based on computed tomography arterial phase images was implemented, utilizing 2.5D deep learning in combination with a MIL framework for the prediction of MVI in HCC. Moreover, this method was compared with the radiomics signature and clinical signatures, and the predictive performance of the various models was evaluated using receiver operating characteristic curves and decision curve analysis (DCA), with DeLong's test applied to compare the area under the curve (AUC) between models. Kaplan-Meier curves were utilized to analyze differences in recurrence-free survival (RFS) or progression-free survival (PFS) among different HCC treatment cohorts stratified by MIL signature risk.
MIL signature demonstrated superior performance in the validation set (AUC = 0.877), significantly surpassing the radiomics signature (AUC = 0.727, = 0.047) and clinical signature (AUC = 0.631, = 0.004). DCA curves indicated that the MIL signature provided a greater clinical net benefit across the full spectrum of risk thresholds. In the prognostic analysis, high- and low-risk groups stratified by the MIL signature exhibited significant differences in RFS within the surgical resection cohort (training set = 0.0058, validation set = 0.031) and PFS within the TACE treatment cohort ( = 0.045).
MIL signature demonstrates more accurate MVI prediction in HCC, surpassing radiomics signature and clinical signature, and offers precise prognostic stratification, thereby providing new technical support for personalized HCC treatment strategies.
微血管侵犯（MVI）是肝细胞癌（HCC）的一个重要预后因素，但其术前预测仍然具有挑战性。
开发并验证一种基于2.5维（2.5D）深度学习的多实例学习（MIL）模型（MIL特征）用于预测HCC中的MVI，评估并将其性能与放射组学特征和临床特征进行比较，并评估其在手术切除和经动脉化疗栓塞（TACE）队列中的预后预测价值。
纳入一个由192例病理确诊的HCC患者组成的回顾性队列，其中68例为MVI阳性，124例为MVI阴性。患者按7:3的比例随机分为训练集（134例患者）和验证集（58例患者）。另外45例接受TACE治疗的HCC患者被纳入TACE验证队列。实施了一种基于计算机断层扫描动脉期图像的建模策略，利用2.5D深度学习结合MIL框架来预测HCC中的MVI。此外，将该方法与放射组学特征和临床特征进行比较，并使用受试者操作特征曲线和决策曲线分析（DCA）评估各种模型的预测性能，应用DeLong检验比较模型之间的曲线下面积（AUC）。利用Kaplan-Meier曲线分析按MIL特征风险分层的不同HCC治疗队列之间无复发生存期（RFS）或无进展生存期（PFS）的差异。
MIL特征在验证集中表现出卓越的性能（AUC = 0.877），显著超过放射组学特征（AUC = 0.727，P = 0.047）和临床特征（AUC = 0.631，P = 0.004）。DCA曲线表明，MIL特征在整个风险阈值范围内提供了更大的临床净效益。在预后分析中，按MIL特征分层的高风险和低风险组在手术切除队列中的RFS（训练集P = 0.0058，验证集P = 0.031）和TACE治疗队列中的PFS（P = 0.045）存在显著差异。
MIL特征在HCC中表现出更准确的MVI预测，超过放射组学特征和临床特征，并提供精确的预后分层，从而为个性化的HCC治疗策略提供新的技术支持。
