表现为磨玻璃结节的孤立性肺结节诊断模型的建立

发布时间：2018-06-22 22:50

本文选题：结节病 + 肺　；参考：《中国医学影像学杂志》2017年06期

【摘要】：目的探讨表现为磨玻璃结节(GGN)的恶性孤立性肺结节(SPN)的独立预测因子,并建立一个预测模型。资料与方法回顾性研究2014年1月—2015年12月上海交通大学附属胸科医院经病理证实的362例(A组)表现为GGN的SPN患者的临床和CT影像特征,筛选出恶性SPN的独立预测因子并建立预测模型。同时收集温州医科大学附属舟山医院119例SPN患者作为B组,用于验证模型诊断效能。结果多因素Logistic回归分析筛选出边界清楚(OR=6.274,P0.01)、边缘光滑(OR=0.391,P0.01)、分叶征(OR=3.387,P0.01)、胸膜牵拉征(OR=2.430,P0.01)及空泡征(OR=3.076,P0.01)为恶性SPN患者的独立预测因子。根据独立预测因子建立的模型受试者工作特性(ROC)曲线下面积为0.859(95%CI:0.804~0.903),诊断准确率85.92%,敏感度91.03%,特异度81.97%,阳性预测值92.03%,阴性预测值73.53%。结论本研究筛选出表现为GGN的恶性SPN的独立预测因子,并建立了可用于准确鉴别SPN的预测模型,可为SPN的早期诊断提供帮助。
[Abstract]:Objective to investigate the independent predictors of malignant solitary pulmonary nodules (SPN) with glass grinding nodules (GGN) and to establish a predictive model. Materials and methods Clinical and CT imaging features of 362 patients (group A) with GGN who were pathologically confirmed in the chest Hospital affiliated to Shanghai Jiaotong University from January 2014 to December 2015 were retrospectively studied. The independent predictive factors of malignant SPN were screened and the prediction model was established. 119 patients with SPN in Zhoushan Hospital affiliated to Wenzhou Medical University were collected as group B to verify the diagnostic efficacy of the model. Results the multivariate logistic regression analysis showed that the borderline was clearly defined (ORG 6.274 P 0.01), the margin was smooth (ORP 0.01), the lobulation sign (ORN 3.387% P0.01), pleural traction sign (ORP 0.01) and vacuole sign (ORN 3.076P0.01) were independent predictors of malignant SPN. The area under the operating characteristic (ROC) curve of model subjects established by independent predictors was 0.859 (95 CI: 0.804 / 0.903), the diagnostic accuracy was 85.92, the sensitivity was 91.03, the specificity was 81.97, the positive predictive value was 92.03, and the negative predictive value was 73.53. Conclusion in this study, independent predictors of malignant SPN with GGN were screened, and a predictive model was established for the accurate identification of SPN, which may be helpful for the early diagnosis of SPN.
【作者单位】：温州医科大学附属舟山医院胸心外科;上海交通大学附属胸科医院胸外科;舟山市肺癌研究中心;温州医科大学附属舟山医院病理诊断中心;温州医科大学附属舟山医院放射诊断中心;
【基金】：国家卫生计生委科学研究基金-浙江省医药卫生重大科技计划(省部共建计划)(WKJ2014-2-021) 浙江省科技厅公益技术社会发展项目(2015C33254) 国家自然科学基金青年基金项目(81301996)
【分类号】：R563;R816.41

【参考文献】