Computer-aided diagnosis (CADx) improves the accuracy of radiologic assessment of lung cancer, according to a study published in Academic Radiology.
Researchers from Johns Hopkins University in Baltimore, MD, sought to evaluate the improved accuracy of lung cancer radiologic assessment if CADx was added to the process. The researchers performed a systematic review of the literature for studies that evaluated CADx for lung cancer with chest X-ray or CT. They found 14 articles, representing 1,868 scans, for inclusion in the study. In nine studies, the accuracy of the radiologist was evaluated twice, once without CADx and then again with CADx. These were evaluated for the review. The five other studies used pathological diagnosis for verification.
Among the nine observer-performance studies, there was an average of 10.3 radiologists per study, and eight of these studies showed a significant accuracy improvement with CADx. “Increases in the receiver operating characteristic (ROC) area under the curve of .8 or higher were seen in all nine studies that reported it, except for one that employed subspecialized radiologists,” the authors wrote.
They concluded that their review demonstrated that CADx improved the accuracy of lung cancer assessment over manual review, suggesting that more research be conducted into incorporating its use into screening and regular clinical workflow.