【摘要】 目的探讨CT图像纹理分析鉴别诊断表现为磨玻璃密度结节的肺腺癌浸润性的价值。方法收集在我院接受肺部CT检查且手术病理证实为肺腺癌患者100例(浸润性腺癌56例,非浸润性腺癌44例)。随机选择69例为训练组,31例为验证组。使用A.K.(Analysis-Kinetics)分析软件进行影像特征提取;Kruskal-Wallis非参数检验和Spearman相关性分析进行特征降维;使用R语言软件包"GLM"函数,建立Logistic回归模型;以交叉验证方法对回归模型进行检验。采用ROC曲线评价独立预测因素的诊断效能。结果影像特征提取得到396个影像组学特征,经降维最终得到与鉴别肺非浸润腺癌与浸润腺癌最相关的参数3个,建模后验证Logistic回归模型示其诊断准确率为83.30%,敏感度及特异度分别为77.80%、91.70%。结论 CT图像纹理分析可有效鉴别表现为磨玻璃密度结节肺腺癌的浸润性。更多还原

【Abstract】 Objective To explore the value of CT texture analysis in differential diagnosing invasion of pulmonary adenocarcinoma manifesting ground glass density nodule. Methods Totally 100 patients with pulmonary adenocarcinoma manifesting ground glass density nodule( 56 invasive adenocarcinomas and 44 non-invasive adenocarcinomas) confirmed by pathology underwent CT scanning. Patients were randomly divided into training group( n = 69) and validation group( n = 31). Image features were extracted using A. K.( Analysis-Kinetics) analysis software,and feature dimensionality reduction was conducted with Kruskal-Wallis and Spearman analysis. The Logistic model was established with R language package " GLM" function,then regression model was tested with cross-validation method. ROC curve analysis was performed to evaluate the differentiating value of identified variables. Results Totally 396 texture parameters were obtained from imaging features,and of which 3 features had the relationship with differential diagnosis of invasion of pulmonary adenocarcinoma manifesting ground glass density nodule. ROC curve showed that area under the curve of the Logistic model in validation group was 83. 30%,and the sensitivity and specificity were 77. 80% and 91. 70%,respectively. Conclusion Texture analysis has the potential to improve the differentiation of invasion of pulmonary adenocarcinoma manifesting ground glass density nodule.更多还原