基于3D CT影像组学特征在非小细胞肺癌隐匿性淋巴结转移预测中的应用  

Application of 3D CT Radiomics Characteristics in Prediction of Occult Lymph Node Metastasis in Non-small Cell Lung Cancer

作　　者：吴明明 陈艾琪 杜小萌 卢楚鸣 王凯 邹文涛 赵以惠 王月燕 李伟 左盼莉 马宜传 WU Ming-ming;CHEN Ai-qi;DU Xiao-meng;LU Chu-ming;WANG Kai;ZOU Wen-tao;ZHAO Yi-hui;WANG Yue-yan;LI Wei;ZUO Pan-li;MA Yi-chuan(Department of Radiology,The First Affiliated Hospital of Bengbu Medical University,Bengbu 233004,Anhui Province,China;School of Graduate,Bengbu Medical University,Bengbu 233030,Anhui Province,China;Department of Respiratory and Critical Care Medicine,The First Affiliated Hospital of Bengbu Medical University,Bengbu 233004,Anhui Province,China;Clinical Research Center for Respiratory Disease(Tumor)in Anhui Province,Bengbu 233004,Anhui Province,China;Huiyi Huiying Medical Technology Co.,LTD.,Beijing 100000,China;Anhui Key Laboratory of Digital Medicine and Smart Health,Bengbu 233004,Anhui Province,China)

机构地区：[1]蚌埠医科大学第一附属医院放射科,安徽蚌埠233004 [2]蚌埠医科大学研究生院,安徽蚌埠233030 [3]蚌埠医科大学第一附属医院呼吸与危重症医学科,安徽蚌埠233004 [4]安徽省呼吸系统疾病(肿瘤)临床医学研究中心,安徽蚌埠233004 [5]汇医慧影医疗科技有限公司,北京100000 [6]数字医学与智慧健康安徽省重点实验室,安徽蚌埠233004

出　　处：《中国CT和MRI杂志》2025年第1期50-52,共3页Chinese Journal of CT and MRI

基　　金：研究生科研创新计划项目(Byycx22138);安徽省中央引导地方科技发展资金项目(2020b07030008);领航菁英科研项目专项基金(XM-HR-YXFN-2021-05-19);安徽省临床医学研究转化专项立项项目(202304295107020072)。

摘　　要：目的建立3D CT影像组学特征模型,探讨其在非小细胞肺癌(non-small cell lung cancer,NSCLC)隐匿性淋巴结转移中的预测价值。方法收集2019年1月-2023年6月蚌埠医学院第一附属医院诊治的151例NSCLC患者的临床及影像学资料。所选取的患者在行根治术前,均行胸部增强CT检查,影像报告均无淋巴结肿大/转移提示,术后病理为有/无淋巴结转移。数据集按照7:3随机分成训练组和测试组,在原发肿瘤显示的层面逐层手动勾画,提取最佳影像组学特征,建立预测模型并验证。通过受试者工作特征ROC曲线下面积AUC评估模型的预测效能。结果训练组AUC值为0.885,准确度为0.870,特异度为0.800,敏感度为0.800;测试组AUC值为0.859,准确度为0.840,特异度为0.780,敏感度为0.750。结论基于原发灶3D CT影像组学特征模型,在预测NSCLC隐匿性淋巴结转方面具有可行性。Objective To establish a 3D CT imaging feature model and explore its application value in predicting occult lymph node metastasis in non-small cell lung ca ncer(NSCLC).Methods Clinical and imaging data of 151 NSCLC patients diagnosed and treated in the First Affiliated Hospital of Bengbu Medical College from January 2019 to June 2023 were collected.Before radical surgery,all the selected patients underwent enhanced CT examination,and the image report showed no indication of lymph node enlargement or metastasis,while postoperative pathology was found with or without lymph node metastasis.The data set was randomly divided into the training group and the test group according to 7:3,manually sketched layer by layer at the level of the primary tumor display,extracted the best image omics features,established the prediction model and verified.The predictive efficiency of the model was evaluated by area AUC under receiver operating characteristic ROC curve.Results In the training group,AUC value was 0.885,accu racy was 0.870,specificity was 0.800,sensitivity was 0.800;the AUC value of the test group was 0.859,the accu racy was 0.840,the specificity was 0.780,and the sensitivity was 0.750.Conclusion It is feasible to predict occult lymph node metastasis in NSCLC based on 3D CT imaging feature model of primary lesion.

关 键 词：NSCLC 隐匿性淋巴结转移 3D勾画 影像组学 

分 类 号：R322.25[医药卫生—人体解剖和组织胚胎学]