Bosniak 2019分级肾囊性病变CT特征与病理的对照分析

Comparative Analysis of CT Imaging Features and Pathology of Bosniak 2019 Classification Renal Cystic Lesions

目的评价Bosniak 2019分级系统CT特征对良恶性肾囊性病变的诊断性能。方法回顾性搜集病理证实的79例肾囊性病变患者的临床及CT影像资料。两名医师采用2019版Bosniak分级系统评价其CT影像学特征,并与病理进行对照分析。采用组内相关系数(ICC)评价测量数据的一致性,采用卡方检验、独立样本t检验及Mann-Whitney U检验比较良恶性病变患者临床信息及CT影像特征的差异,将CT特征绘制受试者工作特征(ROC)曲线,使用Z检验比较CT特征之间的诊断效能。结果良恶性病变间囊壁或强化分隔的最大厚度、分隔数量、是否伴有不规则增厚及结节、是否伴有钙化差异具有统计学意义。在所有病变中,囊壁及分隔最大厚度的曲线下面积(AUC)为0.925,高于单独采用分隔数量,差异具有统计学意义,高于联合分隔数量的诊断效能,但差异无统计学意义。结论Bosniak 2019分级系统更新的定量影像学特征可在一定程度上反映病变的良恶性,在临床上具有良好的应用价值。

Objective To evaluate the diagnostic performance of the Bosniak 2019 classification system CT features for benign and malignant renal cystic lesions.Methods Clinical information and CT images of 79 patients with renal cystic lesions who had pathology were collected retrospectively.The 79 lesions were analyzed by two physicians using the 2019 version of the Bosniak classification system to evaluate their CT imaging features,which were compared to pathological findings.ICC was used to evaluate the interobserver agreement of the measured features.The chi-square test,independent samples t-test,and Mann-Whitney U test were used to compare the differences in clinical information and imaging features of patients with benign and malignant lesions.The ROC curves were plotted for each quantitative feature,and the diagnostic efficacy between the different features was compared using the Z test.Results The differences in the maximum thickness of the wall or septa,the number of septa,the presence of irregular thickening and nodules,and the presence of calcification between the two groups for benign and malignant lesions were statistically significant.In all lesions,the AUC of the maximum thickness of the cyst wall and septa was 0.925,which was statistically higher than using the number of septa alone and higher than the combination of maximum thickness and the number of septa.However,the difference was not statistically significant.Conclusion The updated quantitative imaging features of the Bosniak 2019 classification system can reflect the benignity and malignancy of lesions,which has good clinical application.