基于能谱CT特征的无或少实性成分磨玻璃结节样肺腺癌浸润程度鉴别诊断模型构建  

Differential diagnosis model construction of invasive degree of lung adenocarcinoma manifesting as ground-glass nodules with no or little solid component based on energy spectrum CT features

作　　者：刘毅勇[1] 霍凤芝 Liu Yiyong;Huo Fengzhi(Department of Imaging,969th Hospital of Joint Logistics Support Force of Chinese People's Liberation Army,Hohhot 010051,China;Department of Oncology(Respiratory),Inner Mongolia Hospital,Peking University Cancer Hospital,Hohhot 010020,China)

机构地区：[1]中国人民解放军联勤保障部队第九六九医院影像科,呼和浩特010051 [2]北京大学肿瘤医院内蒙古医院肿瘤内科(呼吸),呼和浩特010020

出　　处：《国际肿瘤学杂志》2025年第4期197-201,共5页Journal of International Oncology

摘　　要：目的构建基于能谱CT特征的无或少实性成分磨玻璃结节(GGN)样肺腺癌浸润程度的鉴别诊断模型,旨在为后续临床诊疗工作提供借鉴。方法回顾性纳入2019年1月—2022年12月于中国人民解放军联勤保障部队第九六九医院和北京大学肿瘤医院内蒙古医院行手术治疗肺腺癌且CT表现为无或少实性成分GGN的患者145例,根据浸润程度分为浸润组(51例)和微浸润组(94例)。采用logistic回归进行无或少实性成分GGN样肺腺癌浸润程度鉴别诊断影响因素的多因素分析,并构建logistic回归预测模型。采用受试者操作特征(ROC)曲线进行各指标的预测效能分析。结果单因素分析结果显示,结节最大径(t=-6.30,P<0.001)、结节平均CT值(t=-5.43,P<0.001)、空气支气管征(χ^(2)=23.21,P<0.001)、微血管CT成像类型(χ^(2)=27.94,P<0.001)均是无或少实性成分GGN样肺腺癌浸润程度鉴别诊断的预测因素。多因素logistic回归分析结果显示,结节最大径(OR=1.72,95%CI为1.33~2.23,P<0.001)、结节平均CT值(OR=1.01,95%CI为1.01~1.02,P<0.001)、空气支气管征(OR=4.92,95%CI为1.59~15.21,P=0.006)及微血管CT成像Ⅲ型(OR=14.01,95%CI为2.97~66.06,P=0.001)均是无或少实性成分GGN样肺腺癌浸润程度鉴别诊断的独立预测因素。利用上述多因素分析的结果构建logistic回归模型:logit(P)=0.54×结节最大径+0.01×结节平均CT值+1.59×空气支气管征+2.64×微血管CT成像类型(Ⅲ型)-3.33。ROC曲线分析显示,结节最大径、结节平均CT值、空气支气管征、微血管CT成像类型、logistic回归模型P值进行无或少实性成分GGN样肺腺癌浸润程度鉴别诊断的曲线下面积分别为0.759、0.751、0.686、0.741、0.918。结论包括结节最大径、结节平均CT值、空气支气管征及微血管CT成像类型在内的能谱CT特征可用于无或少实性成分GGN样肺腺癌浸润程度鉴别诊断;利用以上4个因素构建的logistic回归模型对于患者的浸润程度�Objective To construct the differential diagnosis model of invasive degree of lung adenocarcinoma manifesting as ground-glass nodules(GGNs)with no or little solid component based on energy spectrum CT features,and to provide reference for the follow-up clinical diagnosis and treatment.Methods A retrospective study was conducted on 145 patients who underwent surgical treatment for lung adenocarcinoma at the 969th Hospital of Joint Logistics Support Force of Chinese People's Liberation Army and Inner Mongolia Hospital of Peking University Cancer Hospital from January 2019 to December 2022,presenting with CT findings of no or little solid component GGNs.The patients were divided into invasive group(51 cases)and microinvasive group(94 cases)based on the invasive degree.Logistic regression was used to conduct a multivariate analysis of factors affecting the differential diagnosis of invasive degree of lung adenocarcinoma manifesting as GGNs with no or little solid component,and construct a logistic regression model.Receiver operator characteristic(ROC)curve was used to analyze the predictive efficiency of each index.Results Univariate analysis showed that the maximum diameter of nodules(t=-6.30,P<0.001),average CT value of nodules(t=-5.43,P<0.001),air bronchial sign(χ^(2)=23.21,P<0.001),microvascular CT imaging type(χ^(2)=27.94,P<0.001)were predictors of invasive degree of lung adenocarcinoma manifesting as GGNs with no or little solid component.Multivariate logistic regression analysis showed that the maximum diameter of nodules(OR=1.72,95%CI:1.33-2.23,P<0.001),average CT value of nodules(OR=1.01,95%CI:1.01-1.02,P<0.001),air bronchial sign(OR=4.92,95%CI:1.59-15.21,P=0.006)and microvascular CT imaging typeⅢ(OR=14.01,95%CI:2.97-66.06,P=0.001)were independent predictors of invasive degree of lung adenocarcinoma manifesting as GGNs with no or little solid component.A logistic regression model was constructed based on the results of the above multiple factor analysis:logit(P)=0.54×maximum diameter of the nodule+0.01�

关 键 词：体层摄影术 X线计算机 肺腺癌 肿瘤浸润 诊断 鉴别 磨玻璃结节 

分 类 号：R734.2[医药卫生—肿瘤]