Detectability of chemically induced periapical lesions by limited cone beam computed tomography, intra-oral digital and conventional film radiography

Abstract

Our aim was to compare the Accu-I-Tomo (3DX), the Digora Optime image plate system and F-speed film in detecting chemically created apical lesions. Lesions were created by applying perchloric acid apical to extracted teeth in jaw specimens for 1, 1.5 or 2 h. After being repositioned, teeth were radiographed with Accu-I-Tomo limited cone beam CT (LCBCT), Digora Optime storage phosphor plates (SPP) and F-speed films. Six observers scored the presence of lesions using a five-grade scale. The accuracy of each observer and modality was assessed through receiver operating characteristic (ROC) analysis and A(z) values were compared using two-way ANOVA. Pairwise comparisons of imaging systems were carried out using the Mann-Whitney U-test. Differences in A(z) values were compared using Friedman and Dunn's tests. Kappa (kappa) was used to measure interobserver agreement. The A(z) values were larger for LCBCT than for SPP and film for all acid durations. For 1 h of acid duration a significant difference was found between LCBCT and film (P = 0.02) and between LCBCT and SPP (P = 0.0043), For 1.5 h a significant difference (P = 0.006) was found between LCBCT and SPP only. For 2 h acid duration, there was no significant difference between LCBCT and film or SPP (P>0.05). Between SPP and film no significant difference was found for any acid duration (P>0.05). kappa ranged between fair and moderate for LCBCT and between slight and fair for SPP and film. LCBCT images provided better than or similar detectability as film and SPP images of chemically created periapical lesions.