基于示卓安超声造影Kupffer期的深度学习预测肝细胞癌微血管侵犯的研究  

Deep learning based on Kupffer phase of Sonazoid contrast-enhanced ultrasound to predict microvascular invasion in hepatocellular carcinoma

作　　者：秦琼 文荣 白秀梅 高瑞智 杨院萍 甘翔宇 廖蔚 阙翘 陈玉姬 何云[1] 杨红[1] QIN Qiong;WEN Rong;BAI Xiumei;GAO Ruizhi;YANG Yuanping;GAN Xiangyu;LIAO Wei;QUE Qiao;CHEN Yuji;HE Yun;YANG Hong(Department of Medical Ultrasound,The First Affiliated Hospital of Guangxi Medical University,Nanning 530021,Guangxi Province,China)

机构地区：[1]广西医科大学第一附属医院超声医学科,广西南宁530021

出　　处：《肿瘤影像学》2024年第4期432-441,共10页Oncoradiology

基　　金：广西壮族自治区自然科学基金(2020GXNSFDA238005,2023GXNSFDA026013);南宁市青秀区科技计划(2020045);广西医科大学第一附属医院临床研究攀登计划青年科技启明星计划(YYZS2020024)。

摘　　要：目的:评估基于注射用全氟丁烷微球[商品名示卓安(Sonazoid)]超声造影Kupffer期的深度学习模型预测肝细胞癌(hepatocellular carcinoma,HCC)微血管侵犯(microvascular invasion,MVI)的效能,并将其与影像组学模型及临床模型进行比较。方法:回顾并纳入2020年7月—2022年9月于广西医科大学第一附属医院接受Sonazoid超声造影检查的146例原发性HCC患者,以7∶3随机划分为训练集102例和验证集44例。基于肿瘤感兴趣区,使用ResNet101模型通过迁移学习提取深度学习特征,使用PyRadiomics提取影像组学特征。采用Mann-Whitney U检验、最小绝对收缩和选择算子(least absolute shrinkage and selection operator,LASSO)算法进行特征降维。LASSO回归用于构建深度学习模型和影像组学模型,同时还基于临床特征构建一个临床模型。采用受试者工作特征曲线的曲线下面积(area under the curve,AUC)、灵敏度、特异度和准确度评估模型的诊断效能。DeLong检验用于比较模型间的诊断效能。结果:在训练集中,深度学习模型、影像组学模型、临床模型的AUC(95%CI)分别为0.931(0.880~0.981)、0.823(0.744~0.903)、0.719(0.614~0.824)。在验证集中,深度学习模型、影像组学模型、临床模型的AUC(95%CI)分别为0.895(0.757~1.000)、0.711(0.514~0.909)、0.606(0.390~0.822)。DeLong检验表明在训练集和验证集中,深度学习模型的诊断效能均优于影像组学模型及临床模型(P<0.05)。单因素及多因素logistic回归分析示甲胎蛋白和巴塞罗那临床肝癌分期可作为HCC患者MVI的独立预测因子(P<0.01)。结论:基于Sonazoid超声造影Kupffer期的深度学习模型在预测HCC患者MVI方面表现出优异的性能,有望成为预测MVI的无创影像学生物标志物。Objective:To evaluate the performance of a deep learning model based on the Kupffer phase of perflubutane microspheres for injection(product name Sonazoid)contrast-enhanced ultrasound in predicting microvascular invasion(MVI)of hepatocellular carcinoma(HCC),comparing it with radiomics model and clinical model.Methods:This study retrospective included 146 patients with primary HCC who underwent Sonazoid contrast-enhanced ultrasound examination in The First Affiliated Hospital of Guangxi Medical University from July 2020 to September 2022,randomly divided into a training set of 102 and a validation set of 44 in a 7∶3 ratios.Based on the region of interest in tumors,ResNet101 model was used to extract deep learning features through transfer learning,and PyRadiomics was utilized to extract radiomics features.Mann-Whitney U test and least absolute shrinkage and selection operator(LASSO)algorithm were employed to reduce features dimension.LASSO regression was used to construct both the deep learning model and radiomics model,a clinical model was also built based on clinical features.The diagnostic performance of models was evaluated using the area under the receiver operating characteristic curve(AUC),sensitivity,specificity,and accuracy.DeLong testing algorithm was used to compare the diagnostic performance between models.Results:In the training set,the AUC(95%CI)for the deep learning model,radiomics model,clinical model was 0.931(0.880-0.981),0.823(0.744-0.903)and 0.719(0.614-0.824),respectively.In the validation set,the AUC(95%CI)for the deep learning model,radiomics model,clinical model was 0.895(0.757-1.000),0.711(0.514-0.909)and 0.606(0.390-0.822),respectively.DeLong testing indicated that in both the training and validation sets,the diagnostic performance of the deep learning model was superior to that of the radiomics model and clinical model(P<0.05).Both univariate and multivariate logistic regression analyses showed that AFP(P<0.05)and Barcelona Clinic Liver Cancer staging(P<0.001)could be used as independe

关 键 词：肝细胞癌 超声造影 注射用全氟丁烷微球 Kupffer期 深度学习 微血管侵犯 

分 类 号：R735.7[医药卫生—肿瘤] R445.1[医药卫生—临床医学]