【摘要】 目的 探讨HRCT征象与肺内纯磨玻璃结节(pGGN)恶性度的相关性，并使用Nomogram图建立评估pGGN浸润性风险的量化模型。方法 对我院53例pGGN患者(68个病灶)的影像及病理资料进行回顾性分析，HRCT影像资料包括：病灶形态、大小、“分叶”征、“毛刺”征、“空泡”征、血管“集束”征、胸膜“凹陷”征、瘤肺界面以及支气管“充气”征；依据病理资料将pGGN分为非浸润性病变和浸润性病变。运用卡方检验、独立样本t检验对两组HRCT征象进行单因素分析，依据单因素分析结果进一步行Logistic回归多因素分析，筛选出评估pGGN浸润性的独立风险因素。最后引入R软件(R3.3.2)，使用rms软件包，构建Nomogram图风险评估模型。结果 68个pGGN病灶中，非浸润性病变28例、浸润性病变40例。Logistic回归多因素分析显示，病灶大小、“分叶”征、“空泡”征、“毛刺”征以及血管“集束”征是评估pGGN浸润性的独立风险因素。以此构筑的Nomogram图模型，病灶大小(100分)、“分叶”征(92分)、“毛刺”征(90分)、血管“集束”征(68分)、“空泡”征(66分)，将危险因素的积分相加为总分，总分对应模型评估pGGN浸润性的风险性。通过Nomogram图积分进行受试者工作曲线(Receiver operating curve,ROC)分析时，曲线下面积为0.870，对应的敏感度、特异度分别为87.50%、71.43%。结论 病灶大小、分叶征、空泡征、毛刺征以及血管集束征是评估pGGN浸润性的独立风险因素，本研究构建的Nomogram图模型有助于其量化评估。更多还原

【Abstract】 Objective To investigate the correlation between HRCT signs and malignancy of Pure ground glass nodules(pGGN) in lung,and to establish a quantitative model for evaluating the risk of pGGN infiltration using Nomogram. Methods The imaging and pathological data of 53 patients with pGGN(68 lesions) in our hospital were retrospectively analyzed. HRCT imaging data included lesion shape,size,lobulation sign,burr sign,vacuole sign,vascular cluster sign,pleural indentation sign,tumorlung interface and bronchial inflation sign. According to the pathological result,pGGN was divided into pre-invasive lesions and invasive lesions. Chi-square test and independent sample t-test were used to analyze the HRCT signs of the two groups. Logistic regression multivariate analysis was carried out based on the results of single factor analysis to screen out independent risk factors for evaluating the infiltration of pGGN. Finally,R software(R3.3.2) is introduced to construct Nomogram risk assessment model using RMS software package. Results Among 68 pGGN lesions,28 were pre-invasive lesions and 40 were invasive lesions. Logistic regression analysis showed that lesion size,lobulation sign,vacuole sign,burr sign and vascular cluster sign were independent risk factors for evaluating pGGN infiltration. Nomogram showed lesion size(100 points),lobulation sign(92 points),spiculation sign(90 points),vessel cluster sign(68 points),vacuole sign(66 points). The risk of pGGN infiltration was assessed by the corresponding total score model. The area under the curve was 0.870,and the sensitivity and specificity were 87.50% and 71.43% respectively. Conclusion Focal size,lobulation sign,vacuole sign,spiculation sign and vascular cluster sign are independent risk factors for evaluating the infiltration of pGGN. And this study is helpful for its quantitative evaluation.更多还原