摘要
目的 采用基于多参数磁共振序列的卷积神经网络(convolutional neural network, CNN)联合传统影像组学标签及临床指标,术前预测肝细胞性肝癌(hepatocellular carcinoma, HCC)患者的微血管侵犯(microvascular invasion, MVI)。方法 选择经病理确诊的HCC患者275例纳入本研究。将数据集随机分为训练集(n=192)和测试集(n=83)。应用CNN技术,融合二维多参数磁共振肿瘤图像、三维肿瘤的传统影像组学特征标签及临床指标,开发一种HCC的MVI预测分类器。应用受试者工作特征曲线(receiver operating characteristic, ROC),比较混合模型(Model^(Com))与卷集神经网络模型(Model^(D))、影像组学模型(Model^(R))和临床模型(Model^(C))的诊断效能。结果 Model^(D)在训练集和测试集中的AUC分别为0.914和0.842,优于Model^(C)(训练集:P<0.001;测试集:P=0.032)和Model^(R)(训练集:P<0.001;测试集:P=0.044)。Model^(Com)在训练集和测试集中的AUC分别为0.951和0.881,在训练集中优于Model^(D)(P=0.012),在测试集中差异无统计学意义(P=0.157)。校准曲线显示出了Model^(Com)具有良好的拟合优度(hosmer-lemeshow test,训练集P=0.402,测试集P=0.689)。决策曲线分析提示Model^(Com)鉴别MVI阳性和MVI阴性的净获益高于其他模型。结论 CNN为基础的混合模型够准确预测HCC的MVI状态。
Objective To explore convolutional neural network(CNN)based on multi-parameter magnetic resonance(MR)sequences,combined with traditional radiomics signature and clinical indicators,in predicting microvascular invasion(MVI)of hepatocellular carcinoma(HCC)preoperatively.Methods Two hundred and seventy-five patients with pathologically confirmed HCC were enrolled in this study.The data set was randomly divided into a training set(n=192)and a test set(n=83).A MVI predictive classifier for HCC was developed by using CNN technique,which fused 2D multi-parameter MR tumor images,3D traditional radiomics signature and clinical indicators.Using the receiver operating characteristic(ROC)curve,the performance of combined model(Model^(Com)),CNN model(Model^(D)),radiomics model(Model^(R))and clinical model(Model^(C))were compared.Results The area under the ROC curve(AUC)of Model^(D) was 0.914 in the training set and 0.842 in the test set,which was better than that of Model^(C)(training set:P<0.001;test set:P=0.032)and Model^(R)(training set:P<0.001;test set:P=0.044).The AUC of Model^(Com) in the training set and test set were 0.951 and 0.881,respectively,which was better than that of Model^(D) in the training set(P=0.012),but there was no significant difference in the test set(P=0.157).The calibration curve showed that Model^(Com) had a good goodness of fit(Hosmer-Lemeshow test,P=0.402 for training set,P=0.689 for test set).Decision curve analysis showed that the net benefit of Model^(Com) in identifying positive MVI and negative MVI was higher than that of other models.Conclusion The Model^(Com) based on CNN can accurately predict the MVI status of HCC.
作者
王谦
王会哲
卢双动
沈丹平
刘龙艳
WANG Qian;WANG Hui-zhe;LU Shuang-dong;SHEN Dan-ping;LIU Long-yan(Department of Radiology,the Second Central Hospital of Baoding City,Hebei Province,Zhuozhou 072750,China;Department of Gastroenterology,the Second Central Hospital of Baoding City,Hebei Province,Zhuozhou 072750,China;Department of Emergency,the Second Central Hospital of Baoding City,Hebei Province,Zhuozhou 072750,China;Department of CT,the Second Central Hospital of Baoding City,Hebei Province,Zhuozhou 072750,China;Department of Otorhinolaryngology,the Second Central Hospital of Baoding City,Hebei Province,Zhuozhou 072750,China)
出处
《河北医科大学学报》
CAS
2024年第7期771-778,共8页
Journal of Hebei Medical University
基金
保定市科技计划项目(2341ZF037)。
关键词
肝肿瘤
卷积神经网络
影像组学
liver neoplasms
convolutional neural network
radiomics
