Diagnostic accuracy of cone beam computed tomography in detection of simulated mandibular condyle erosions

To determine the diagnostic accuracy of cone beam computed tomography (CBCT) in the detection of simulated mandibular condyle erosions. Seventeen dry human mandibles were used in this in vitro study. NewTom VG CBCT scanner (New Tom VG, Verona, Veneto region, Italy) was used for the condyles imaging (pre-erosion and post-erosion image). Thirty three lesions were created on the superior (11 cases), anterior (11 cases), and posterior surfaces (11 cases) of the condyles. The pre- and post-erosion images were randomly presented to two previously calibrated oral and maxillofacial radiologists in order to evaluate the presence of simulated erosions and their position in the condyles using two protocols. In the first protocol, axial and coronal images and in the second protocol, axial, coronal, and sagittal/multiplanar reconstructed (MPR) images were used to evaluate the lesions of the samples. Furthermore, the Cochran''s Q test and McNemar and Kappa statistical tests were used to assess the sensitivity, specificity, and accuracy of this study. There was no statistically significant difference between the diagnostic methods and the reference value. There was substantial agreement between the two protocols (Kappa > 0.61). Protocol 2 showed relatively better results than protocol 1 but the difference was not statistically significant (P > 0.05). Sensitivity, specificity, and diagnostic accuracy levels in the erosion imaging were higher in the posterior region of condyle; however, there was no statistically significant difference between the condylar regions (P > 0.05). CBCT had high sensitivity, specificity, and diagnostic accuracy in the detection of simulated mandibular condyle erosions.