Effects of dental material-induced artifacts on signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) in CBCT: an ex vivo study

This ex vivo study aimed to evaluate the material-dependent effects of a high metal artifact reduction (MAR) algorithm on quantitative cone-beam computed tomography (CBCT) image-quality parameters assessed using signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Nine dental material conditions—an amalgam restoration, composite restoration, glass ionomer cement restoration, metal crown, orthodontic bracket, root canal filling material, calcium hydroxide [Ca(OH)₂] medicament, implant body, and implant abutment—together with one intact control tooth were evaluated in ten separate dry human mandibles. Each condition was represented by a single specimen. A total of ten CBCT acquisitions were performed using a single CBCT unit and a standardized acquisition protocol. Each acquisition was reconstructed without MAR and with high MAR, resulting in 20 reconstructed CBCT image datasets. Image analysis was performed using ImageJ2/Fiji. SNR and CNR were calculated from standardized regions of interest placed adjacent to the evaluated materials without including the materials themselves. The significance level was set at p < 0.05. MAR application produced statistically significant changes in the SNR adjacent to most materials ( p < 0.001). SNR values generally increased with high MAR, whereas a significant reduction was observed adjacent to the metal crown ( p < 0.001). No significant SNR change was detected adjacent to the root canal filling material ( p > 0.05). High MAR significantly increased CNR values adjacent to the metal crown, orthodontic bracket, composite restoration, and implant abutment ( p < 0.001), whereas CNR significantly decreased adjacent to the amalgam restoration and implant body ( p < 0.05) and the glass ionomer cement restoration ( p = 0.047). No significant CNR differences were observed for the root canal filling material, calcium hydroxide [Ca(OH)₂] medicament, or control tooth ( p > 0.05). Within the limitations of this exploratory ex vivo study, the effect of high MAR on quantitative image-quality parameters, as assessed using SNR and CNR, varied according to the type of dental material. Because each material condition was represented by a single specimen, the findings should be interpreted cautiously. The results are specific to the CBCT unit, acquisition protocol, and high-MAR setting used in this study. Further studies including multiple independent specimens and observer-based diagnostic assessments are required to determine whether these quantitative changes influence diagnostic performance.