肺实性结节半自动体积测量提高观察者间Lung-RADS评分一致性

Semi-automatic volumetry of solid pulmonary nodules improves the interobserver consistency of Lung-RADS score

摘要:

目的 研究3D-Slicer软件实性肺结节体积测量对不同观察者在肺部影像报告数据系统（lung CT screening reporting and data system，Lung-RADS）分类一致性中的影响。方法 纳入76例患者中的76个实性结节。由3位放射科医师分别采用手动和3D-Slicer软件半自动体积测量方法获得结节的直径与体积，并转化为相应的Lung-RADS评分，其中2分为阴性，3分及以上为阳性。采用同类相关系数（intraclass correlation coefficient，ICC）及Bland-Altman指数来评价观察者间直径与体积测量的一致性，Kappa分析评价观察者间Lung-RADS评分及阳性/阴性组间的一致性。结果 ICC分析结果显示手动直径测量的一致性（0.994~0.996）明显低于半自动体积测量的一致性（0.997~0.998），同时Bland-Altman指数分析结果显示手动直径测量的偏倚高于半自动体积测量。采用半自动体积测量，能够比手动直径测量明显提高观察者间Lung-RADS评分及阳性/阴性之间的一致性（0.963~0.975及0.957~0.977 vs. 0.833~0.866及0.863~0.892）。结论 3D-Slicer半自动体积测量能够提高实性肺结节观察者间大小测量的一致性，相应的Lung-RADS分类一致性也随之提高。

Abstract:

Objective To study the effect of solid pulmonary nodules volume measurement by 3D-Slicer software on the interobserver consistency of lung CT screening reporting and data system（Lung-RADS） categories.Methods Seventy-six solid nodules from 76 patients were included in this study. The diameter and volume of nodules were obtained by three radiologists using manual and 3D-Slicer software semi-automatic volume measurement methods，respectively，and the raw data was converted to the corresponding Lung-RADS score，of which 2 points was negative and 3 points or above was positive. Intraclass correlation coefficient（ICC） and Bland-Altman index were used to evaluate the interobserver consistency of diameter and volume measurements. Kappa analysis was used to evaluate the interobserver consistency of Lung-RADS score and the intergroup consistency of positive/negative.Results ICC analysis showed that the consistency of manual diameter measurement（0.994-0.996） was significantly lower than that of semi-automatic volume measurement（0.997-0.998），and Bland-Altman index analysis showed that the bias of manual diameter measurement was higher than that of semi-automatic volume measurement. Compared with manual diameter measurement，the use of semi-automatic volume measurement can significantly improve the consistency of Lung-RADS score and positive/negative classifications（0.963-0.975 and 0.957-0.977 vs. 0.833-0.866 and 0.863-0.892）.Conclusion 3D-Slicer semi-automatic volume measurement can improve the interobserver consistency of the size measurement of solid pulmonary nodules， as well as the corresponding Lung-RADS categories.