RESUMO
INTRODUCTION: The purpose of the present study was (1) to test the accuracy of a small-volume cone-beam computed tomographic (CBCT) device in detecting horizontal root fractures (HRFs) in teeth with and without an intracanal metallic post (IMP) and (2) to investigate the use of 2 different acquisition protocols of a CBCT device for HRF diagnosis. METHODS: Forty endodontically treated teeth with and without an IMP were examined using PreXion 3D CBCT scanner (Terarecon, San Mateo, CA) with a 5-cm high and 5-cm diameter cylinder at 0.10-mm voxel reconstruction. Two observers analyzed the samples to determine the presence and location of HRFs. RESULTS: Sensitivity values ranged from 0.40-0.80. The most favorable results were found for the samples with no IMP observed using the protocol of a higher number of x-ray projections (0.70-0.80). Accuracy in the groups with an IMP ranged from 75%-90% in the 1024 x-ray projection protocol (HI-HI group) versus 70%-85% for the same samples examined in the 512 x-ray projection protocol (HI-STD group). Intraobserver agreement ranged from relevant to perfect concordance for both protocols (HI-HI = Kappa: 0.60-1.00 and HI-STD = Kappa: 0.55-0.89). Interobserver agreement ranged from moderate to perfect concordance for both protocols (HI-HI = Kappa: 0.79-0.89 and HI-STD = Kappa: 0.42-0.76). CONCLUSIONS: Even though there are statistically significant differences for the protocol with the higher number of x-ray projections, we found high accuracy, sensitivity, sensibility, and intra- and interobserver agreement in detecting HRFs for both Prexion 3D protocols.