The Performance of a Zirconium-based Root Filling Material with Artifact Reduction Properties in the Detection of Artificially Induced Root Fractures Using Cone-beam Computed Tomographic Imaging

• Published in: Journal of endodontics Vol. 44; no. 5; pp. 828 - 833
• Main Authors: Fox, Aaron, Basrani, Bettina, Lam, Ernest W.N.
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01-05-2018
• Subjects: Beam hardening artifact; cone-beam computed tomography; Dentistry; gutta-percha; vertical root fracture; zirconium; cone-beam computed tomography; Beam hardening artifact; vertical root fracture; zirconium; gutta-percha
• ISSN: 0099-2399; 1878-3554
• DOI: 10.1016/j.joen.2018.02.007

Limited field of view cone-beam computed tomographic (CBCT) imaging has been used to augment clinical testing of vertical root fractures (VRFs); however, the presence of gutta-percha (GP) in the canal space generates substantial imaging artifacts that make fracture detection difficult. The purpose of this study was to evaluate the influence of a zirconium (Zr)-based root filling material with radiologic properties that reduce beam hardening (BH) artifacts using CBCT imaging in the in vitro diagnosis of VRFs.
One hundred seventy-six single-rooted mandibular premolar teeth were obtained, and half of these teeth were filled with GP or Zr (CPoint; EndoTechnologies, LLC, Shrewsbury, MA). VRFs were induced in 44 decoronated teeth in each group using an Instron (Norwood, MA) Universal Testing Machine. Each root was then placed in a dry human mandible and imaged with the Carestream 9000 3D CBCT system (Carestream Dental, Atlanta, GA). The images were evaluated by 6 oral maxillofacial radiologists (OMRs) and residents.
The sensitivity was greater for detecting VRFs in the Zr group than the GP group (P = .035). However, the specificity was greater for the GP group than the Zr group (P = .028). Receiver operating characteristic area under the curve values were greater for the Zr group than the GP group, but these differences were not statistically significant. The OMRs outperformed the residents in the detection of VRFs in the Zr group with respect to specificity (P = .006) and positive predictive value (P = .012).
The reduced BH of the Zr group improved the sensitivity of the detection of artificially induced VRFs. The ability to detect VRFs in the Zr group was further enhanced by clinical experience.