摘要
目的 基于双层探测器光谱CT平扫定量参数联合形态学特征,在与胸膜接触的纯磨玻璃结节(pure ground-glass nodules with pleural contact,P-pGGNs)中鉴别浸润性腺癌(invasive adenocarcinoma,IAC)和微浸润腺癌(microinvasive adenocarcinoma,MIA).方法 回顾性搜集术前行双层探测器光谱CT胸部平扫,表现为P-pGGNs且手术病理证实为MIA和IAC患者104例(111个P-pGGNs),分为IAC组(n=33)和MIA组(n=78).测量结节长径、短径及三维CT值、电子云密度和有效原子序数值.临床资料包括性别和年龄;形态学特征包括结节形状、边界及分叶、毛刺、胸膜牵拉、空泡、空气支气管和血管异常征.比较IAC与MIA组形态及定量参数,筛选独立预测因素;受试者工作特征(ROC)曲线评价及比较各独立预测因素以及联合预测模型的诊断效能.结果 与MIA组相比,IAC组结节多表现不规则、边界模糊,且更易出现分叶、毛刺、空泡、空气支气管和血管异常征(P＜0.05).IAC组年龄、结节长径、短径、三维CT值和ED值均高于MIA组(P＜0.05).Logistic回归分析结节三维CT值和ED值、长径、形状、分叶征及空泡征是IAC的独立预测因子.联合预测模型受试者工作特征曲线曲线下面积(AUC)高达0.953,灵敏度90.91％,特异度89.74％.结论 双层探测器光谱CT平扫多定量参数联合形态学特征构建预测模型鉴别P-pGGNs的IAC和MIA具有较高准确性.
Abstract
Objective To explore the clinical application of dual-layer spectral detector CT plain scan for differentiating the invasive adenocarcinoma from minimally invasive adenocarcinoma in pure ground-glass nodules with pleural contact(P-pGGNs).Methods 104 patients with total 111 P-pGGNs,who underwent preoperative dual-layer spectral detector CT scan were included retrospectively.According to invasiveness by pathology,the nodules were divided into two groups:inva-sive adenocarcinoma(IAC,n=33)group and minimally invasive adenocarcinoma(MIA,n=78)group.The electron cloud density map and effective atomic number map were reconstructed on dedicated workstation(IntelliSpace Discovery),and the maximum diameter and vertical short diameter,as well as the 3D CT value,electron density value(ED value)and effec-tive atomic number(Zeffvalue)of P-pGGNs were measured.The clinical data included sex and age.The qualitative param-eters included P-pGGN shape,interface,lobulation,spiculation,pleural retraction,bubble-like lucency,air bronchogram and vascular abnormality.Diagnostic performance was evaluated by receiver operating characteristic(ROC)curve analysis.Results Significant differences were observed from nodule shape,interface,lobulation,spiculation,bubble-like lucency,air bronchogram,vascular abnormality,age,maximum diameter and vertical short diamete,3D CT value and ED value between IAC and MIA groups(P＜0.05).Logistic regression analysis revealed that the 3D CT value[odds ratio(OR)=0.0919,P=0.010],3D ED value(OR=3.384,P=0.003),maximum diameter(OR=2.747,P=0.014),shape(OR=17.740,P=0.018),lobulation(OR=0.030,P=0.012)and bubble-like lucency(OR=0.029,P=0.006)were inde-pendent factors for diagnosing I AC manifesting as P-pGGNs.The analysis of ROC curves showed that the AUC of the com-bination of six independent predictors was 0.953(sensitivity=90.91％,specificity=89.74％).Conclusion The com-bination of qualitative and quantitative parameters of dual-layer spectral detector CT plain scan displayed a noteworthy ca-pacity to differentiate IAC from MIA in P-pGGNs.
基金项目
国家自然基金项目(81960310)
云南省临床医学研究中心子课题(202102AA100067)
NETL
NSTL
万方数据