基于临床及CT影像组学特征构建周围型小细胞肺癌与肺腺癌诊断模型的研究  

Study on the Differential Diagnosis Model between Peripheral SCLC and ADC Based on Clinical and CT Radiomics Features

作　　者：汪靖婷 钟飞扬 甘甜 龙云 廖美焱[1] WANG Jingting;ZHONG Feiyang;GAN Tian(Department of Radiology,Zhongnan Hospital of Wuhan University,Wuhan,Hubei Province 430071,P.R.China)

机构地区：[1]武汉大学中南医院放射科,430071

出　　处：《临床放射学杂志》2023年第3期406-410,共5页Journal of Clinical Radiology

摘　　要：目的基于临床及CT影像组学特征建立周围型小细胞肺癌(SCLC)与肺腺癌(ADC)诊断模型,并评估其诊断价值。方法回顾性搜集周围型肺癌患者临床及CT影像资料,选取治疗前2周内有薄层CT影像的病例分为SCLC组和ADC组,以SCLC组为实验组,采用倾向性评分匹配按1∶2匹配ADC对照组,两组按照7∶3比例随机分为训练集和验证集。依据训练集病例资料采用多因素Logistic回归分析筛选有意义的变量,建立临床、影像组学及临床组学联合预测周围型SCLC诊断模型。采用受试者工作特征(ROC)曲线评价模型诊断效能,建立个体化诊断列线图。结果周围型SCLC和ADC两组间NSE和13个组学特征有显著性差异。训练集和验证集ROC曲线下面积,临床诊断模型分别为0.793和0.750,影像组学模型分别为0.857和0.838,联合模型分别为0.905和0.882。结论基于临床及CT影像组学特征建立周围型SCLC与ADC诊断模型可鉴别诊断周围型SCLC及ADC。Objective To evaluate the value of differential diagnosis model of peripheral small cell lung cancer(SCLC)and lung adenocarcinoma(ADC)based on clinical and CT radiomics features.Methods The clinical and CT image data of patients with peripheral lung cancer were retrospectively analyzed.The patients with HRCT images within 2 weeks before treatment were divided into SCLC group and ADC group.SCLC group was used as the experimental group,and the PSM was used to match the ADC control group according to 1∶2.The two groups were randomly divided into training set and verification set in a 7∶3 ratio.According to the data of the training set,multivariate logistic regression analysis was used to screen meaningful variables,and the clinical,radiomics and combined prediction models were established respectively.The receiver operating characteristic curve(ROC)was used to evaluate the diagnostic efficiency of the models,and an individualized diagnostic nomogram was established.Results There were significant differences in NSE and 13 selected radiomics characteristics between peripheral SCLC and ADC groups.The area under ROC curve(AUC)of training set and validation set were 0.793 and 0.750 for clinical diagnosis model,0.857 and 0.838 for imaging group model and 0.905 and 0.882 for combined model respectively.Conclusion The diagnostic model of peripheral SCLC and ADC based on clinical and CT radiomics features shows good diagnostic performance for the classification of peripheral SCLC and ADC.

关 键 词：小细胞肺癌 肺腺癌 体层摄影术 X线计算机 影像组学 诊断模型 

分 类 号：R734.2[医药卫生—肿瘤]