Computer-aided diagnosis of pulmonary nodules: Results of a large-scale observer test

PURPOSE: To determine the effect of computer-aided diagnosis (CAD) on the a ccuracy of pulmonary nodule detection. MATERIALS AND METHODS: Twenty abnormal chest radiographs, each with a singl e nodule, and 20 normal radiographs were digitized with a laser scanner. Th ese images were analyzed by using a computer program that indicates areas t hat may represent pulmonary nodules. The radiographs were displayed on comp uter workstations in randomized order, and an observer test was performed. One hundred forty-six observers participated, including 23 chest radiologis ts, 54 other radiologists, 27 radiology residents, and 42 nonradiologists. Cases were interpreted first without and then with the use of CAD. The obse rvers' responses were recorded on a continuous confidence rating scale. Det ection accuracy both with and without CAD was evaluated with receiver opera ting characteristic analysis. RESULTS: The detection accuracy was significantly higher for all categories of observers when CAD was used (chest radiologists, P = 8 x 10(-6) other r adiologists, P = 2 x 10(-16); radiology residents, P = 6 x 10(-7); and nonr adiologists, P = 8 x 10(-9)). CONCLUSION: CAD has the potential to improve diagnostic accuracy in the det ection of lung nodules on digital radiographs.