Jin A, et al. - Through an analysis of a reading map diagnosis competition, experts assessed the accuracy of human and computer-aided diagnosis (CAD) for the discovery of thyroid cancers via ultrasonography. Thirty-three thyroid nodule images randomly selected between 2015 and 2017 were included. In the competition, 177 participants including one operator using CAD were involved. The competition had an online segment and the live segment. The accuracy of contestants was 60.3% while and the accuracy of the CAD system was 84.8%. With various technical titles, the accuracy of the CAD system was greater than that of the contestants. For CAD and contestants, the areas under the curve were 0.985 and 0.659, respectively. Hence, the CAD system was observed as having great accuracy in this thyroid nodule reading map competition.