Impact of Voxel-Size in CBCT Dental Implant Measurements in Vitro

Objectives: This study aimed to evaluate accuracy and error of dental implant CBCT-images comparing diverse CBCT-voxel size in-vitro.
Methods: A radiographically visible block was made for this study. Ten implants (diameter: 3.5mm; length: 11mm) were placed perpendicular to the block, using a jig, exactly to the block level. Two canals simulating the mandibular canal were made into the block in a distance 1mm and 2mm from the implant apex, 14 and 15mm from the block surface, respectively (5 implants per group). The block was scanned using a Kodak 9000 CBCT machine at the recommended manufacturer’s exposure settings (70kVp, 10mAs) and at 0.076, 0.1 and 0.2mm voxel-sizes. The scans were analyzed in the CS3D-imaging program. Two blinded examiners measured implant length, diameter and distance between implant apices and the canal using various voxel sizes. Statistical comparison (differences) of measurements with the real measurements using different voxel sizes was performed with a linear mixed model. Significant difference level was set p ≤ 0.05.
Results: The data showed that there is no statistically significant difference among voxel-sizes for all measurements (Table).
Conclusions: Acceptable measurement accuracy and minimal error was shown in all CBCT-images, regardless of voxel-size.