How do imaging protocols affect the assessment of root-end fillings?

Objectives: This study investigated the impact of micro-computed tomography (micro-CT)-based voxel size on the analysis of material/dentin interface voids and thickness of different endodontic cements. Materials and Methods: Following root-end resection and apical preparation, maxillary premolars were filled with mineral trioxide aggregate (MTA), Biodentine, and intermediate restorative material (IRM) (n = 24). The samples were scanned using micro- CT (SkyScan 1272; Bruker) and the cement/dentin interface and thickness of materials were evaluated at voxel sizes of 5, 10, and 20 mu m. Analysis of variance and the Tukey test were conducted, and the degree of agreement between different voxel sizes was evaluated using the Bland and Altman method (p < 0.05). Results: All materials showed an increase in thickness from 5 to 10 and 20 mu m (p < 0.05). When evaluating the interface voids, materials were similar at 5 mu m (p > 0.05), while at 10 and 20 mu m Biodentine showed the lowest percentage of voids (p < 0.05). A decrease in the interface voids was observed for MTA and IRM at 20 mu m, while Biodentine showed differences among all voxel sizes (p < 0.05). The Bland-Altman plots for comparisons among voxel sizes showed the largest deviations when comparing images between 5 and 20 mu m. Conclusions: Voxel size had an impact on the micro-CT evaluation of thickness and interface voids of endodontic materials. All cements exhibited an increase in thickness and a decrease in the void percentage as the voxel size increased, especially when evaluating images at 20 mu m.