Qin Qiong, Pang Jinshu, Li Jingdan, Gao Ruizhi, Wen Rong, Wu Yuquan, Liang Li, Que Qiao, Liu Changwen, Peng Jinbo, Lv Yun, He Yun, Lin Peng, Yang Hong
Department of Ultrasound, The First Affiliated Hospital of Guangxi Medical, University, Nanning, Guangxi, China.
Department of Ultrasound, Fujian Medical University Union Hospital, Fuzhou, China.
Med Phys. 2025 Jul;52(7):e17895. doi: 10.1002/mp.17895. Epub 2025 May 19.
Microvascular invasion (MVI) is strongly associated with the prognosis of patients with hepatocellular carcinoma (HCC).
To evaluate the value of Transformer models with Sonazoid contrast-enhanced ultrasound (CEUS) in the preoperative prediction of MVI.
This retrospective study included 164 HCC patients. Deep learning features and radiomic features were extracted from arterial and Kupffer phase images, alongside the collection of clinicopathological parameters. Normality was assessed using the Shapiro-Wilk test. The Mann‒Whitney U-test and least absolute shrinkage and selection operator algorithm were applied to screen features. Transformer, radiomic, and clinical prediction models for MVI were constructed with logistic regression. Repeated random splits followed a 7:3 ratio, with model performance evaluated over 50 iterations. The area under the receiver operating characteristic curve (AUC, 95% confidence interval [CI]), sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV), decision curve, and calibration curve were used to evaluate the performance of the models. The DeLong test was applied to compare performance between models. The Bonferroni method was used to control type I error rates arising from multiple comparisons. A two-sided p-value of < 0.05 was considered statistically significant.
In the training set, the diagnostic performance of the arterial-phase Transformer (AT) and Kupffer-phase Transformer (KT) models were better than that of the radiomic and clinical (Clin) models (p < 0.0001). In the validation set, both the AT and KT models outperformed the radiomic and Clin models in terms of diagnostic performance (p < 0.05). The AUC (95% CI) for the AT model was 0.821 (0.72-0.925) with an accuracy of 80.0%, and the KT model was 0.859 (0.766-0.977) with an accuracy of 70.0%. Logistic regression analysis indicated that tumor size (p = 0.016) and alpha-fetoprotein (AFP) (p = 0.046) were independent predictors of MVI.
Transformer models using Sonazoid CEUS have potential for effectively identifying MVI-positive patients preoperatively.
微血管侵犯（MVI）与肝细胞癌（HCC）患者的预后密切相关。
评估基于索纳造影剂增强超声（CEUS）的Transformer模型在术前预测MVI中的价值。
本回顾性研究纳入了164例HCC患者。从动脉期和 Kupffer 期图像中提取深度学习特征和影像组学特征，并收集临床病理参数。使用 Shapiro-Wilk 检验评估数据正态性。应用 Mann-Whitney U 检验和最小绝对收缩和选择算子算法筛选特征。采用逻辑回归构建MVI的Transformer、影像组学和临床预测模型。按照7:3的比例进行重复随机分割，在50次迭代中评估模型性能。使用受试者工作特征曲线下面积（AUC，95%置信区间[CI]）、敏感性、特异性、准确性、阳性预测值（PPV）、阴性预测值（NPV）、决策曲线和校准曲线评估模型性能。应用 DeLong 检验比较模型之间的性能。采用 Bonferroni 方法控制多重比较产生的 I 型错误率。双侧 p 值<0.05 被认为具有统计学意义。
在训练集中，动脉期Transformer（AT）模型和 Kupffer 期Transformer（KT）模型的诊断性能优于影像组学和临床（Clin）模型（p<0.0001）。在验证集中，AT和KT模型在诊断性能方面均优于影像组学和Clin模型（p<0.05）。AT模型的AUC（95%CI）为0.821（0.7—0.925），准确率为80.0%；KT模型的AUC（95%CI）为0.8——0.977），准确率为70.0%。逻辑回归分析表明，肿瘤大小（p=0.016）和甲胎蛋白（AFP）（p=0.046）是MVI的独立预测因素。
基于索纳造影剂CEUS的Transformer模型在术前有效识别MVI阳性患者方面具有潜力。
