肺孤立结节CT征象及定量参数预测早期浸润性腺癌病理亚型的价值  

The Value of CT Features and Quantitative Parameters of Solitary Pulmonary Nodules in Predicting Pathological Subtypes of Early Invasive Adenocarcinoma

作　　者：尹雪梅 李文菲[1] 李晓超[2] 李英杰[1] 张俊 Yin Xue-Mei;Li Wen-Fei;Li Xiao-Chao;Li Ying-Jie;Zhang Jun(Department of Medical Imaging,The First Hospital of Qinhuangdao,QinHuangDao 066000,HeBei Province,China;Department of Pathology,The First Hospital of Qinhuangdao 066000,HeBei Province,China)

机构地区：[1]河北省秦皇岛市第一医院医学影像科,河北秦皇岛066000 [2]河北省秦皇岛市第一医院病理科,河北秦皇岛066000

出　　处：《中国CT和MRI杂志》2023年第8期53-56,共4页Chinese Journal of CT and MRI

基　　金：秦皇岛市科技局计划项目(202004A089)。

摘　　要：目的分析肺孤立结节CT征象及定量参数,基于逻辑回归模型预测早期浸润性腺癌(IAC)病理亚型的价值。方法回顾性分析最大径<3cm的238例肺孤立结节早期IAC患者不同病理亚型分级的临床及影像数据;比较病理亚型各级间CT特征及定量参数是否存在统计学差异,经多元Logistic回归分析确定独立危险因素并构建预测中高危病变组的二元Logistic回归模型,绘制ROC曲线,确定最佳鉴别诊断效能。结果(1)平均直径、体积、密度直方图CT平均值、异常CT支气管征、肿瘤微血管CT成像征、分叶征、毛刺征在鉴别三组间存在统计学差异(P<0.05)。(2)密度直方图CT平均值、异常CT支气管征是鉴别三组间的独立危险因素(P均<0.05),构建的联合二元Logistic回归模型的鉴别诊断效能最高(AUC=0.92)。结论密度直方图CT平均值、异常CT支气管征是鉴别肺孤立结节早期IAC病理亚型不同分级的重要CT特征,联合二元Logistic回归模型对于孤立结节早期IAC中高危病变有较高的诊断效能。Objective To analyze the CT signs and quantitative parameters of solitary pulmonary nodules,and predict the value of pathological subtypes of early invasive adenocarcinoma(IAC)based on logistic regression model.Methods Retrospectively analysed of clinical and imaging data of different pathological subtypes in 238 patients with early-stage IAC with solitary pulmonary nodules<3cm.Comparing the CT features and quantitative parameters of different pathological subtypes,whether there are statistical difference,The independent risk factors were determined by multiple Logistic regression analysis,and the binary Logistic regression model for predicting the moderate and high risk lesion groups was constructed,and the ROC curve was drawn to determine the best differential diagnosis efficiency.Results(1)The average diameter,volume,density histogram CT average value,abnormal CT bronchus sign,tumor microvascular CT imaging sign,lobulation sign,and spur sign had statistically significant differences among the three groups(P<0.05).(2)Density histogram CT average value and abnormal CT bronchus sign were independent risk factors for distinguishing among the three groups(all P<0.05).The ROC curve showed that the constructed joint binary Logistic regression model had the highest differential diagnosis efficiency(AU C=0.92).Conclusion Density histogram CT mean value and abnormal CT bronchus sign are important CT features to distinguish different grades of pathological subtypes of early IAC in solitary pulmonary nodules.The combined binary Logistic regression model has a high diagnostic efficiency for early IAC lesions with high risk in solitary nodules.

关 键 词：结节 肺腺癌 CT征象 定量参数 病理亚型 

分 类 号：R563[医药卫生—呼吸系统]