Background: The major use of cone-beam computed tomography (CBCT) is in implant planning. Although CBCT has been shown to be a reliable tool for measurements, we were unable to find studies comparing the differences between images acquired using different voxel sizes for implant planning. The aims of the present in vitro study are to verify the accuracy of linear measurements and to analyze the visibility of the mandibular canal on CBCT images obtained using different voxel sizes.
Methods: Twelve dry human hemimandibles were scanned using voxel sizes of 0.2, 0.3, and 0.4 mm, and 108 transversal CBCT images were generated, on which two examiners performed 648 linear measurements and evaluated the visibility of the mandibular canal.
Results: The global mean ± SD error between measurements on images and direct measurements was 0.23 ± 0.20 mm. CBCT measurements underestimated direct measurements in 390 cases (60.2%). All measurement errors found were <1 mm. There was a statistically significant difference (P = 0.0001 and P = 0.0028) between the two examiners only when measuring the width in two posterior sites. There was a statistically significant difference (P = 0.0242) between voxel sizes only for one of the examiners, for a measurement of height. It was possible to visualize the mandibular canal in all of the 108 images evaluated.
Conclusions: Transversal CBCT images are adequate for linear measurements in the posterior region of the mandible and provide adequate visualization of the mandibular canal using voxel sizes of 0.2, 0.3, or 0.4 mm. A voxel of 0.3 mm is a good compromise between image quality and low radiation dose.