基于CT影像组学评估胃腺癌人表皮生长因子受体-2表达状态的价值  

The value of CT-based radiomics for evaluating the expression status of human epidermal growth factor receptor 2 in gastric adenocarcinoma

作　　者：王素雅[1] 詹鹏超 邢静静 梁盼[1] 岳松伟[1] 张永高[1] 高剑波[1] WANG Su-ya;ZHAN Peng-chao;XING Jing-jing(Department of Radiology,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450052,China)

机构地区：[1]郑州大学第一附属医院放射科,郑州450052

出　　处：《放射学实践》2025年第1期9-15,共7页Radiologic Practice

摘　　要：目的:基于CT影像组学及临床特征建立可无创性评估胃腺癌人表皮生长因子受体2(HER-2)表达状态的预测模型,并验证其效能。方法:回顾性分析本院263例经病理确诊为胃腺癌患者的临床资料,其中HER-2阳性患者49例,阴性214例。将所有患者以7:3的比例随机分为训练集(n=185)和验证集(n=78)。基于3D Slicer软件和门脉期CT图像手动勾画病灶感兴趣区(ROI),并提取影像组学特征。在训练集中,比较HER-2阳性与阴性组患者的临床特征差异,采用多因素Logistic回归确定临床独立预测因子,建立临床模型。基于最小绝对收缩和选择算子(LASSO)回归算法构建影像组学模型,计算影像组学分数(Radscore)。结合临床独立预测因子和Radscore构建联合模型。根据ROC曲线的曲线下面积(AUC)评估模型的预测效能,绘制校准曲线评价模型预测概率与真实概率之间的一致性,采用决策曲线(DCA)分析模型的临床价值。结果:肿瘤厚径(OR=1.04,P=0.033)、cT分期(OR=2.39,P=0.038)、cN分期(OR=2.15,P=0.046)为HER-2阳性表达的临床独立预测因子。在训练集中,临床模型、影像组学模型和联合模型预测胃癌HER-2阳性表达的的AUC分别为0.711(0.626~0.795)、0.852(0.787~0.917)和0.872(0.808~0.936);在验证集中,临床模型、影像组学模型和联合模型预测胃癌HER-2阳性表达的的AUC分别为0.698(0.534~0.861)、0.818(0.698~0.938)和0.853 (0.747~0.959)。校准曲线显示联合模型预测概率与真实概率之间的一致性良好,DCA结果显示联合模型可为胃腺癌患者提供临床净获益。结论:基于CT影像组学和临床特征构建的联合模型可用于治疗前无创性评估胃腺癌的HER-2表达状态。Objective:The aim of this study is to develop and validate a non-invasive prediction model for evaluating the expression status of human epidermal growth factor receptor 2(HER-2)in gastric adenocarcinoma.This prediction model is based on CT radiomic features and clinical characteristics.Methods:In this study,a retrospective analysis was conducted on 263 clinical data of patients diagnosed with gastric adenocarcinoma at our hospital,out of which 49 cases were HER-2 positive and 214 cases were HER-2 negative.Patients were randomly divided into a training set(n=185)and a validation set(n=78)at a ratio of 7:3.Radiomic features were extracted by manually delineating the region of interest(ROI)on venous phase CT images using 3D slicer software.In the training set,clinical features differences between HER-2 positive and negative groups were compared.Independent clinical predictors were determined using multivariate logistic regression and were utilized to develop a clinical model.A radiomics model was also created using a least absolute shrinkage and selection operator(LASSO)algorithm,which calculated the radiomics score(Radscore)for each patient.An integrated model that combined the independent clinical predictors and Radscore was also constructed as a final model.The predictive performance of the model was evaluated using the area under the receiver operating characteristic curve(AUC),while calibration curves were used to assess the agreement between predicted and observed probabilities.Additionally,decision curve analysis(DCA)was conducted to examine the clinical utility of the models.Results:Tumor thickness(OR=1.04,P=0.033),cT stage(OR=2.39,P=0.038),and cN stage(OR=2.15,P=0.046)were identified as the independent clinical predictors for HER-2 positive expression.In the training set,the area under the receiver operating characteristic curve(AUC)values of the clinical,radiomics,and combined models were 0.711(0.626~0.795),0.852(0.787~0.917),and 0.872(0.808~0.936),respectively.Similarly,in the validation set,the respecti

关 键 词：胃肿瘤 人表皮生长因子受体-2 表达状态 体层摄影术 X线计算机 影像组学 

分 类 号：R735.2[医药卫生—肿瘤] R814.42[医药卫生—临床医学]