中国临床医学影像杂志2025，Vol.36Issue(3)：168-173,6.DOI:10.12117/jccmi.2025.03.004
基于Logistic回归探讨CT形态学参数对高级别肺腺癌的预测价值
Exploring the predictive value of CT morphological parameters for high-grade lung adenocarcinoma based on Logistic regression
摘要
Abstract
Objective:To explore the predictive value of CT morphological parameters for high-grade lung adenocarcinoma using Logistic regression.Methods:A total of 187 patients who were diagnosed with lung adenocarcinoma and underwent rad-ical surgical resection of lung cancer in our hospital from January 2022 to September 2023 were retrospectively selected as the study objects.According to the pathological results,they were divided into high-grade lung adenocarcinoma group(n=73)and non-high-grade lung adenocarcinoma group(n=114).Clinical and pathological data as well as CT morphological parameters(including long/short diameter of lesion,average CT value,lobulation,burr,vacuole,shape,density,air bronchogram and pleu-ral traction sign)were recorded and compared between the two groups.Multivariate Logistic regression was used to analyze the independent predictors of high-grade lung adenocarcinoma.ROC curve was used to evaluate the value of CT morphological parameters in the diagnosis of high-grade lung adenocarcinoma.Results:There were no significant differences in gender,age and other aspects between the high-grade lung adenocarcinoma group and the non-high-grade lung adenocarcinoma group.The main histological patterns of the two groups were significantly different(P<0.05).The long diameter,short diameter and average CT value of lesions in the high-grade lung adenocarcinoma group were higher than those of non-high-grade lung adenocarcinoma group.The frequency of lobulation,burr,irregular shape,solid nodules and pleural traction signs in high-grade lung adenocarcinoma group was higher than that in non-high-grade lung adenocarcinoma group(P<0.05).There were no significant differences in cavitation and air bronchial sign between the two groups(P>0.05).The results of multivariate Logistic regression analysis showed that long/short diameter of lesion,average CT value,lobulation,burr,shape,density and pleural traction sign were independent predictors of high-grade lung adenocarcinoma(P<0.05).The AUC values of the above CT mor-phological parameters were 0.714,0.622,0.703,0.705,0.612,0.664,0.594,respectively.The AUC of the combined prediction was 0.944,and the sensitivity and specificity were 90.4%and 86.8%,which was higher than the diagnostic efficiency of the single prediction.Conclusion:CT morphological parameters have good diagnostic value for high-grade lung adenocarcinoma.关键词
肺肿瘤/体层摄影术,X线计算机Key words
Lung Neoplasms/Tomography,X-Ray Computed分类
医药卫生引用本文复制引用
程保国,李卫星,栗鸿宝,李鹏政,秦栓梅..基于Logistic回归探讨CT形态学参数对高级别肺腺癌的预测价值[J].中国临床医学影像杂志,2025,36(3):168-173,6.基金项目
河南省医学科技攻关计划项目(LHGJ20230879). （LHGJ20230879）
