Artifacts caused by titanium implants in CBCT images of the mandible: an experimental study

Abstract

This study evaluated the quantity of metal artifacts produced by dental implants placed in different mandibular regions using various cone beam computed tomography (CBCT) protocols. Titanium implants were placed in 4 regions (incisor, canine, premolar, and molar) of an artificial mandible and subjected to CBCT examinations with the mandibular model placed in different positions within the field of view (FOV) and imaged with different FOV and voxel sizes. An axial section of the cervical region of each implant was selected for artifact quantification. The artifacts were measured by normalizing the actual standard deviation (ASD) of the voxel values. Kruskal-Wallis and Student-Newman-Keuls tests were used to compare the tooth regions and the different positions of the mandible. The Wilcoxon test was used to compare changes in FOV and voxel size. The intraobserver agreement was calculated using the intraclass correlation coefficient. The significance level was 5%. The incisor region showed significantly more artifacts than other regions (P = 0.0315). No statistically significant difference was found when the position of the mandible varied within the FOV (P = 0.7418). Smaller FOV and smaller voxels produced more artifacts (P < 0.0001). The quantity of metal artifacts was affected by FOV and voxel size as well as by anatomical region. Variation of the mandible location within the FOV did not affect the artifacts as defined by the normalized ASD of the voxel values.