Do voxel size and filter application on CBCT images improve the diagnosis of condylar morphological alterations? An ex vivo study

Background There is no established protocol for the low doses of cone‐beam computed tomography (CBCT) acquisition and filter application on the diagnosis of condylar morphological alterations. Objectives The objective of the study was to evaluate the influence of voxel size and filter application on the diagnosis of condylar morphological alterations in CBCT using an ex vivo model. Methods CBCT scans of 36 temporomandibular joints were acquired using OP300 Maxio with voxel sizes of 0.085, 0.125 and 0.280 mm. Three radiologists evaluated the condyles in the CBCT volumes under three filter modes: no filter, ‘1x’ sharpen filter and ‘2x’ sharpen filter. The area under the ROC curve, sensitivity and specificity were calculated by comparing the evaluators' responses with the reference standard and compared among experimental groups using analysis of variance (ANOVA) (α = 5%). Results The area under the ROC curve, sensitivity and specificity were not affected by voxel size and filter application (p > .05). For osteophyte and flattening, there were more true‐positive responses in smaller voxel size. For erosion, the increase of true‐ and false‐positive responses occurred with a larger voxel size. Overall, there was a tendency to decrease true positives for osteophyte and erosion and increase false positives for flattening with the ‘2x’ sharpen filter. Conclusions The diagnosis of condylar morphological alterations assessed in this ex vivo analysis is not influenced by the voxel size and the application of the filters. However, attention is needed with the occurrence of false‐positive diagnosis with a larger voxel size for erosion and ‘2x’ sharpen filter for flattening. Graphical summary of the study's main results.