基于影像组学的矽肺病CT影像分期预测研究  

A study on CT image staging prediction of silicosis based on radiomics

作　　者：杨柳琼 郑建军 向遥 YANG Liuqiong;ZHENG Jianjun;XIANG Yao(Medical School,Ningbo University,Ningbo,Zhejiang 315211,China;Department of Radiology,Hua Mei Hospital,University of Chinese Academy of Sciences,Ningbo,Zhejiang 315010,China)

机构地区：[1]宁波大学医学院,浙江宁波315211 [2]中国科学院大学宁波华美医院放射科,浙江宁波315010

出　　处：《影像研究与医学应用》2023年第3期47-50,共4页Journal of Imaging Research and Medical Applications

摘　　要：目的:探讨基于CT影像组学模型对具有明确矽尘接触史患者进行筛查和分期预测的可行性。方法:选取2016年1月—2020年12月于中国科学院大学宁波华美医院做职业病诊断且具有明确矽尘接触史的患者共245例(0期104例,Ⅰ期77例,Ⅱ期30例,Ⅲ期24例),按照7:3比例随机分为训练组和测试组。分析患者的临床特征及组学特征,采用Logistic回归构建矽肺分期的四分类预测模型,通过受试者工作特征(receiver operating characteristic,ROC)曲线评价模型的预测能力。结果:选取7个组学特征构建影像组学模型,在验证集中0~Ⅲ期矽肺的曲线下面积(area under the curve,AUC)分别为0.86、0.84、0,81、0.95。结论:基于胸部CT图像所构建的影像组学模型对矽肺分期表现出较高的预测效能。Objective To investigate the feasibility of screening and staging prediction based on CT imaging histology model for patients with a clear history of silica dust exposure.Methods A total of 245 patients with occupational disease diagnosis and a clear history of silica dust exposure were selected from January 2016 to December 2020 at Ningbo Huamei Hospital,University of Chinese Academy of Sciences(104 cases in stage 0,77 cases in stageⅠ,30 cases in stageⅡand 24 cases in stageⅢ),and randomly divided into a training group and a test group according to a 7:3 ratio.The clinical and histological characteristics of the patients were analyzed,and a four-category prediction model for silicosis stage was constructed using logistic regression,and the predictive ability of the model was evaluated by the receiver operating characteristic(ROC)curve.Results Seven features were selected in the training set to construct the imaging histology model,and for silicosis staging,the AUCs of the above machine learning classifier in the validation set for stages 0 toⅢsilicosis were 0.86,0.84,0.81 and 0.95.Conclusion The constructed radiomics model based on chest CT images exhibits high predictive efficacy for silicosis staging.

关 键 词：矽肺 CT 影像组学 分期 

分 类 号：R445.3[医药卫生—影像医学与核医学]