Influence of Metal Artifact Reduction Algorithms on the Detection of Inferior Alveolar Canal Perforations by Zirconia Implants in CBCT Images

INTRODUCTION: This study aims to evaluate the efficacy of a metal artifact reduction (MAR) algorithm in detecting contact between zirconia implants and inferior alveolar canal. MATERIALS AND METHODS: Twelve zirconia implants were placed in dry human mandibles using surgical guides, in two conditions: 1 mm above and within the canal. cone beam computed tomography (CBCT) was acquired with the MAR algorithm turned off and on. Three observers indicated whether canal damage was present using a 5-point Likert scale: 1. Definitely no contact, 2. Probably no contact, 3. Uncertain, 4. Probably contact, and 5. Definitely contact. Sensitivity and specificity were calculated and the ROC curve was used to compare the two conditions (α = 0.05). RESULTS: The area under curve (AUC) was higher in MAR On condition for all observers; however, the differences in AUC between the On and Off MAR states were not statistically significant (p > 0.05): for observer 1, AUC increased from 0.9916 (Off) to 1.0 (On); for Observer 2, it increased from 0.7666 to 0.8833; and for Observer 3, it increased from 0.95 to 1.0. Observers 1 and 3 displayed 100% sensitivity and specificity regardless of MAR activation, while for Observer 2, diagnostic sensitivity increased from 66.7% to 83.3% with MAR activation, while specificity remained constant at 80% in both conditions. CONCLUSION: For the detection of contact of the zirconia implants with the IAC, activating the MAR algorithm slightly increased the diagnostic performance metrics of CBCT images, but the differences were not statistically significant. CLINICAL RELEVANCE: Activating the MAR algorithm did not produce a statistically significant improvement in detecting contact between zirconia implants and the inferior alveolar canal.