A novel three-dimensional quantitative assessment method for abnormal root morphology of the maxillary premolars in vivo on cone-beam computed tomography

Abstract

Abstract Background Previous studies have described and recorded abnormal root morphology; however, most of these studies were based on two-dimensional periapical or panoramic radiographs, and only a few studies have quantified it. We aimed to combine two-dimensional periapical radiographs and three-dimensional cone-beam computed tomography (CBCT) to conduct qualitative judgments and quantitative analyses of normal and conical roots, and explore the clinical diagnostic method of normal and conical roots based on intraoral radiographs and CBCT. Methods The conical root was identified visually on periapical radiographs as the clinical gold standard. All teeth were divided into the cone-rooted teeth (CRT) or normal-rooted teeth (NRT) groups. Furthermore, differences in root length (RL), root surface area (RSA), and root volume (RV) of conical and normal roots in the maxillary premolars on CBCT were compared. Receiver operator characteristic curves were generated, and the area under the curve (AUC) and cut-off values were calculated to evaluate the diagnostic value of RV, RSA, RV/RL, and RSA/RL. Results The RSAs of NRT and CRT were 236.88 ± 27.93 mm2 and 207.98 ± 27.80 mm2, respectively (P = 0.000). The mean RV in the CRT group was lower than that in the NRT group, and the difference was statistically significant (253.40 ± 41.98 mm3 vs. 316.93 ± 49.89 mm3, P = 0.000). The RSA and RV of conical roots in single root premolars were 12.29% and 19.33% less than those of normal roots, respectively. The AUC values of RSA/RL and RV/RL were 0.87 and 0.89, respectively, and the best cut-off values were 19.61 for RSA/RL (if RSA/RL was < 19.61, the teeth were considered CRT) and 24.05 for RV/RL (if RV/RL was < 24.05, the teeth were considered CRT). Conclusions CBCT has significant diagnostic value in the clinical evaluation of conical roots. RSA/RL and RV/RL were the best parameters with the largest AUC and high sensitivity and specificity.