肺腺癌病灶实性成分的CT三维定量分析对其病理等级的预测价值  

The predictive value of 3D quantitative volumetric analysis of the solid component of lung adenocarcinoma for its pathological grade

作　　者：涂文婷 范丽[1] 顾亚峰 李琼[1] 望云[1] 夏艺[1] 管宇[1] 刘士远[1] TU Wenting;FAN Li;GU Yafeng;LI Qiong;WANG Yun;XIA Yi;GUAN Yu;LIU Shiyuan(Department of Radiology,Changzheng Hospital,Second Military Medical University,Shanghai 200003,China)

机构地区：[1]第二军医大学长征医院影像科,上海200003

出　　处：《实用放射学杂志》2018年第10期1512-1516,共5页Journal of Practical Radiology

基　　金：国家重点研发计划政府间合作项目（2016YFE0103000）;国家自然基金重点项目（81230030）;国家自然基金面上项目（81370035）;国家自然基金重点项目（81401408）;上海市浦江人才计划项目（15PJD002）.

摘　　要：目的 通过对混杂磨玻璃结节（mGGN）实性成分的定量测量，探讨CT三维定量分析在不同病理等级肺腺癌中的预测价值。方法 回顾性分析病理证实为肺腺癌的280例mGGN［不典型腺瘤样增生（AAH）/原位癌（AIS）40例、微浸润性腺癌（MIA）42例、浸润性腺癌（IAC）198例］。采用计算机辅助软件对mGGN进行半自动逐层分割，然后将实性成分的边界勾勒出来，软件可自动生成实性成分的阈值、平均CT值、体积、体积百分比以及通过计算可得到实性成分的质量。3组间差异性比较采用Kruskal-Wallis H检验；将AAH、AIS、MIA纳入A组，IAC纳入B组，2组间差异性比较采用Mann-Whitney U检验，将组别作为因变量、有差异的定量参数作为自变量进行Logistic回归分析。结果 实性成分的阈值、平均CT值、体积、体积百分比、质量在不同病理等级间的差异有统计学意义（P〈0.001），重新分组后也有显著统计学差异（P〈0.001）。Logistic回归分析显示，实性成分的平均CT值［优势比（OR）=1.008，95%可信区间（95%CI）：1.001~1.016，P=0.035］、体积（OR=1.021，95%CI：1.010~1.031，P〈0.001）和质量（OR=0.980，95%CI：0.971~0.990,P〈0.001）是诊断IAC的重要预测因子，对回归预测概率进行受试者工作特征曲线（ROC曲线）分析后，曲线下面积（AUC）为 0.893（P〈0.001）。单独定量参数的ROC曲线显示，平均CT值、体积和质量的最佳诊断界值分别为-202.5 HU，209.0 mm3，168.2 mg。结论 CT三维定量分析可用于对mGGN中实性成分的定量测量，实性成分的平均CT值、体积和质量是诊断mGGN病理等级的重要预测因子。Objective To explore the value of 3D quantitative volumetric analysis in predicting pathological grade of lung adenocarcinoma by quantitative measuring solid component of pulmonary mixed ground glass nodule（mGGN）.Methods 280 mGGN that pathologically confirmed as lung adenocarcinoma 140 cases of atypical adenomatous hyperplasia （AAH）/adenocarcinomas in situ（AIS） ,42 cases of minimally invasive adenocarcinoma （MIA） and 198 cases of invasive adenocarcinoma （IAC）-] were analyzed retrospectively.The border of mC;GN was semi-automatically depicted slice by slice with eomputer-aided software, then the boundary of solid component was outlined.Finally, the software automatically generated threshold, mean CT value, volume, volume percentage of solid component and mass of solid compotlent was calculated.The difference between the three groups was compared by Kruskal-Wallis H test.AAH, AIS, MIA were included in group A, and IAC was included in group B. The difference between the two groups was compared with Mann Whitney U test. Logistic regression analysis was performed by taking group as the dependent variable, and statistically different quantitative parameters were taken as independent variables.Results The threshold, mean CT value jvolume,volume percentage and mass of solid component were statistically different among different pathological grades （P〈0.001）,and above five quantitative parameters between group A and group B were also statistically different after regrouping （P〈0.001 ）.Logistic regression analysis showed that mean CT value（OR=1.008,95%CI：1.001 1.016,P =0.035） ,volume（OR= 1.021,95 %CI： 1.010- 1.031, P〈0.001） and mass（OR=0.980,95%CI：0.971-0.990,P〈0.001 ） of solid component were important predictors that predicting mGGN was IAC.The ROC analysis was performed based on the predicted probability of Logistic regression model, and the AUC was 0.893（P〈0.001）.The ROC curve of individual quantitative parameter showed optimal diagnostic threshold of mean CT v

关 键 词：肺腺癌 磨玻璃结节 实性成分 定量分析 计算机体层成像 

分 类 号：R734.2[医药卫生—肿瘤] R814.42[医药卫生—临床医学]