Abstract
BACKGROUND: Cone Beam Computed Tomography (CBCT) is used to evaluate the hard tissue of the maxillofacial area, and the dose received by the patient depends on the resolution of the images. This study investigated the effect of CBCT resolution (voxel size) on its linear measurement accuracy. METHODS: This in vitro study was conducted on 19 human mandibles and 11 dry maxillae. On each jaw, four anterior and four posterior regions were examined. Markers were placed in the designated areas using gutta-percha. The jaws were scanned at three voxel sizes (100, 150, 200 µm). The bone height and thickness were measured in the prepared images and compared with the sizes obtained. Paired t-test, t-test, one-way ANOVA, and Tukey's post-hoc tests were used for statistical analysis. RESULTS: Buccal and lingual bone height measurements showed significant differences across voxel sizes (p < 0.05), with higher resolution (100 µm) yielding greater accuracy. No significant difference was found in bone thickness measurements between resolutions (p = 0.20). Height measurements were significantly less accurate in anterior than posterior regions (p < 0.05). No significant difference was observed between maxilla and mandible except for buccal height at medium and standard resolutions. All CBCT measurements underestimated actual dimensions. Thickness measurements were more accurate than height measurements. CONCLUSION: Using high resolution in CBCT machines is recommended to reduce linear measurement error when there is a short distance to the critical anatomical structures.