Metal artifact reduction in maxillofacial bone imaging with dental restorations: efficacy of the MRI pointwise encoding time reduction with radial acquisition sequence versus CT.

Abstract

Pointwise Encoding Time Reduction with Radial Acquisition (PETRA) is an ultra-short echo time (UTE) MRI sequence that effectively reduces metal artifacts. Although UTE-MRI is useful for maxillofacial bone imaging, its metal artifact reduction (MAR) effectiveness has not been directly compared with computed tomography (CT) in the same patient cohort. This study evaluated the MAR effectiveness of routine CT (without MAR algorithms) and the PETRA sequence in bone assessments of the maxilla and mandible in patients with dental restorations. Thirty-one patients with dental restorations underwent both routine CT and PETRA. Two observers independently assessed metal artifact severity using a four-point visual scale (0-3) across five anatomical planes: two in the maxilla, two in the mandible, and the occlusal plane (OP). Median scores, inter-observer agreement (weighted kappa), and the relationship between the number of dental restorations and artifact severity were analyzed. Overall, artifact scores for the PETRA sequence tended to be lower than those for routine CT. At the OP, PETRA scores were significantly lower than routine CT for both observers (p < 0.01), yielding scores of 0 or 1 in over 80% of cases. For PETRA, no strong correlation was found between the number of restorations and artifact severity, consistently showing minimal artifacts regardless of the restoration count. PETRA was more effective than routine CT in reducing metal artifacts for maxillofacial bone assessment, particularly at the OP. PETRA's superior bone visualization suggests its potential as a valuable complementary tool for evaluating jawbone conditions.