作者: Min Wu , Etta Pisano , Yuanshui Zheng
DOI: 10.17705/1CAIS.01616
关键词: Predictive value 、 Data mining 、 Receiver operating characteristic 、 Raw data 、 Experimental testing 、 Sensitivity (control systems) 、 Computer science 、 Health care 、 Set (abstract data type) 、 Information system
摘要: This paper describes how to use Receiver Operator Characteristic (ROC) analysis evaluate radiologists’ performance of interpreting digital mammograms in real-time. We developed an experimental testing system, which implemented a set clinical lesion-matching rules prepare raw ROC data. The system can automatically provide detailed evaluations the performance, such as sensitivity, specificity, positive predictive value, negative diagnostic accuracy, curve, and area under curve (Az). Based on preliminary evaluation we found that is appropriate for real-time computer application, directly using data from database, radiology residents.