Effect of observer instruction on ROC study of chest images.
Effect of observer instruction on ROC study of chest images.
复制标题
观察者指导对胸部图像 ROC 研究的影响。
DOI：
10.1097/00004424-199003000-00004
复制
发表时间：
1990
影响因子：
6.7
通讯作者：
Metz,CE
中科院分区：
文献类型：
--
作者：
Gur,D;Rockette,HE;Good,WF;Slasky,BS;Cooperstein,LA;Straub,WH;Obuchowski,NA;Metz,CE
在过去的十年中，接收器工作特性(ROC)分析在许多医学成像应用中得到了应用。为了确保读者信心评级是可分析的(分布良好，以满足曲线拟合算法的收敛要求)和有意义的(限制数据外推)，许多调查人员专门为此目的培训读者。没有关于这种训练对ROC研究结果的可能影响的实验数据。我们进行了一项多观察者、多疾病的研究，其中300张胸部图像在接受培训前后由四位放射科医生进行评分，以提供均匀分布的置信度评分。结果表明，对于我们的数据集，对于间质性疾病和气胸的训练过程，阅读器和疾病特定的准确性没有显著影响。然而，对于模块的检测，两个阅读器的准确性受到显著影响(P>；0.05)，对于正常和异常图像的分类，一个阅读器的总体准确性受到显著影响(P>；0.01)。因此，尽管在免费阅读环境下进行ROC研究存在困难，但在进行ROC研究之前，应仔细考虑任何干预措施对结果的可能影响。
Receiver-operating characteristic (ROC) analysis has been used in many medical imaging applications during the past decade. In order to ensure that reader-confidence ratings are analyzable (well distributed to meet convergence requirements of curve-fitting algorithms) and meaningful (limit extrapolation of the data), many investigators train readers specifically for this purpose. No experimental data are available concerning the possible effects of such training on the results of ROC studies. We performed a multi-observer, multi-disease study in which 300 chest images were rated by four radiologists before and after they were trained to provide well-distributed confidence ratings. The results indicate that for our data set, reader and disease-specific accuracy was not significantly affected by the training process for interstitial disease and pneumothoraces. However, the accuracy of two readers was significantly affected for the detection of modules (P> 0.05), and the overall accuracy of one reader was significantly affected for the classification of normal versus abnormal images (P> 0.01). Thus, in spite of the difficulties associated with the performance of ROC studies in a free-reading environment, one should carefully consider the possible effects of any intervention on the results prior to conducting ROC studies.