Assessment of artefacts produced by metal posts on CBCT images.

AIM

To evaluate artefact intensity in cone beam computed tomography (CBCT) images of two alloys used in metal posts scanned using different exposure parameters.

METHODOLOGY

The sample consisted of 20 single-rooted teeth divided into two groups for use with either a NiCr post or AgPd post. All teeth were scanned with and without their corresponding metal posts and with and without the presence of an extra restored tooth in the arch. The samples were scanned using CS 9000 3D scanner with two exposure protocols: 85 kV 6.3 mA and 85 kV 10 mA. Voxel size and FOV were fixed at 0.100 mm and 5 cm × 3.75 cm. The presence of artefacts was assessed qualitatively by two calibrated observers using the CBCT volume and paired 2D images, and quantitatively by one trained observer, using ImageJ software. Wilcoxon's signed rank, Mann-Whitney, kappa and chi-square tests were used for qualitative analyses. Two-way anova and Tukey's tests were used for quantitative analyses. All analyses were conducted considering the 95% confidence level (α < 0.05).

RESULTS

For the CBCT volume qualitative analysis, significant differences were observed between the metal alloys in the presence of an extra restored tooth, with higher artefact intensity for AgPd when assessing hypodense halos and lines (P = 0.006). Images with two restored teeth had significantly more hypodense and hyperdense lines (P = 0.033). When evaluating exposure parameters and number of restored teeth, the paired image quality analysis revealed significant disagreement between observers for diagnostic image quality (P = 0.001). Quantitative artefact analysis revealed higher artefact intensity for the AgPd posts in the presence of two restored teeth.

CONCLUSION

Although the exposure parameters tested did not interfere with artefact intensity, post alloys with a higher atomic number and the presence of another metal structure in the arch increased artefact intensity and impaired the diagnostic quality of CBCT images.