EFFECT OF A COMPUTER-AIDED DIAGNOSIS SCHEME ON RADIOLOGISTS PERFORMANCE IN DETECTION OF LUNG NODULES ON RADIOGRAPHS

PURPOSE: To evaluate the effect of a computer-aided diagnosis (CAD) sc heme on radiologists' performance in the detection of lung nodules, an d to examine a new method of receiver operating characteristic (ROC) a nalysis. MATERIALS AND METHODS: One hundred twenty radiographs (60 nor mal and 60 abnormal with lung nodules of varying subtlety) were used. Sixteen radiologists (two thoracic, six general, and eight residents) participated in an observer study in which they read both conventional radiographs and digitized radiographs. The radiologists' performance was evaluated with ROC analysis with two different methods (independen t testing and sequential testing) and a continuous rating scale. RESUL TS: A(z) (area under the best fit binormal ROC curve when it is plotte d in the unit square) values obtained from ROC analysis with and witho ut CAD output were 0.940 and 0.894, respectively, in the independent t est and 0.948 and 0.906, respectively, in the sequential test. Finding s with both methods indicated that the CAD scheme statistically signif icantly improved diagnostic accuracy, particularly for radiologists wi th less experience (P < .001). Reading time was not increased when CAD was used. CONCLUSION: The CAD scheme can assist radiologists in the d etection of lung nodules on chest radiographs.