Impact of CBCT acquisition protocols and artifact reduction filters on diagnosing vertical root fractures.

Abstract

To evaluate the impact of acquisition protocols and artifact reduction filters in cone beam computed tomography on diagnosing vertical root fractures in endodontically treated teeth with and without intraradicular posts. We analyzed 480 tomographic images acquired from two J. Morita scanners (0.125- and 0.08-mm voxel sizes protocols), with application of a blooming artifact reduction filter. Three evaluators assessed these images for root fractures using a 5-point Likert scale. Diagnostic accuracy between filters and protocols was determined using generalized linear models with binomial distribution for the outcome, considering protocol, filter, and dental status. Sensitivity, specificity, positive predictive value, and negative predictive value were also estimated for the filters and protocols. The 0.08-mm voxel size protocol demonstrated a significantly higher percentage of accurate diagnosis compared to the 0.125-mm protocol (p = .001). No statistically significant differences (p ≥ .087) were observed for filter application, interaction between protocol and filter, or dental status. Accuracy, sensitivity, and specificity values were respectively: .93, .87, 1.00 (protocol 1); .99, .99, .99 (protocol 2); .98, .96, .99 (no filter); .95, .90, 1.00 (with filter). The new findings found for the two J Morita scanners used in our study were that images acquired using the voxel size of 0.08 mm showed an improvement in the diagnosis of root fractures and the filters in these devices have no relevance significant for the diagnosis.