Accuracy of bone graft loss assessment around titanium and zirconium implants at varying voxel sizes – An ex vivo CBCT study

Cone beam computed tomography (CBCT) may be used for assessing peri-implant bone graft. However, imaging parameters such as voxel size can impact the final image quality and, consequently, diagnostic accuracy. This study aimed to investigate the influence of voxel size on diagnostic accuracy of bone graft loss adjacent to titanium (Ti) and zirconium (Zr) implants using two different CBCT units. Twelve Ti and twelve Zr implants were placed in twelve edentulous dry mandibles. Gaps around the implants were filled with particulate bone graft. In half of the sample, the graft was removed on the buccal side. Images were acquired using two CBCT units (OP300 and Eagle 3D) with different voxel sizes (from 0.2 ​mm to 0.08 ​mm) and evaluated by five observers. Diagnostic values were calculated for each CBCT unit and compared between implant types and voxel sizes, using the Mann-Whitney U test and the Kruskal-Wallis test with a 5 ​% significance level. The diagnostic values for Zr implants were lower compared to Ti (p ​< ​0.05), regardless of voxel size, except for specificity (p ​≥ ​0.05). Voxel size did not significantly influence the diagnostic values (p ​≥ ​0.05) in images from the Eagle 3D unit. For the OP300, AUROC values were higher for the smallest voxel compared to the largest voxel in Zr implants (p ​< ​0.05). Diagnostic accuracy of bone graft loss is rather low around Zr implants, and smaller voxel sizes are recommended, particularly for the OP300 CBCT unit. High diagnostic accuracy is expected for bone grafting assessment around Ti implants regardless of voxel size and CBCT unit.