Low-dose cone-beam computed tomography in simulated condylar erosion detection: a diagnostic accuracy study.

Abstract

The purpose of the present study was to assess the diagnostic accuracy of low-dose cone-beam computed tomography (CBCT) in the detection of simulated mandibular condyle erosions. 102 simulated erosions were performed on the condyles of eight dry human mandibles. Each mandible was subjected to four CBCT scan protocols: high-definition (HD), normal definition (NORM), ultra-low-dose high-definition (ULD-HD), and ultra-low-dose normal definition (ULD-NORM). All scans were analyzed by two observers. The inter-observer and intra-observer agreement as well as the agreement with the gold standard were assessed. The sensitivity, specificity, positive-predictive value, negative-predictive value and accuracy of erosion detection were calculated. A substantial to almost perfect agreement with the gold standard was found regarding the HD protocol and substantial agreement in NORM and ULD-HD protocols; however, moderate agreement was found regarding the ULD-NORM protocol. The sensitivity, specificity and accuracy values were highest for the HD protocol followed by the NORM and ULD-HD which showed comparable results; while, the ULD-NORM protocol showed the least values. The studied ULD-HD CBCT protocol can be recommended for the detection of mandibular condylar erosions due to the reduced radiation dose; however, ULD-NORM is not advocated for similar clinical use.