Imaging of occlusal dentine caries: a comparison among conventional radiographs, digital radiographs, and cone-beam computed tomography images

Abstract

To compare conventional radiographs, digital radiographs, equalized images of digital radiographs, and cone-beam computed tomography (CBCT) images for in vitro imaging of occlusal dentine caries. A total of 160 extracted human posterior teeth were evaluated in the study. All teeth were free of restorations and proximal caries. The clinical appearances of the occlusal surfaces of the teeth ranged from sound to discolored, with no macroscopic cavitation. Conventional intraoral radiographs, digital intraoral radiographs, and CBCT images of the teeth were taken. The images were evaluated independently by three observers for the presence of occlusal dentine caries. The teeth were then sectioned and computerized images were obtained for the histological sections using a digital camera linked to a stereomicroscope at 10× magnification. The histological images were assessed for the presence or absence of dentinal caries. The sensitivity, specificity, accuracy, and area under the receiver-operating characteristic curve (Az value) were calculated for each diagnostic method. Sensitivity was greatest for CBCT images (mean = 0.63) and lowest for digital radiographs (mean = 0.38). Specificity was greatest for digital radiographs (mean = 0.76) and lowest for CBCT images (mean = 0.64). The mean Az values ranged from 0.56 to 0.60 (P > 0.05). The interobserver and intraobserver agreements were greatest for CBCT images (mean kappa = 0.66 and 0.73, respectively). There were no differences in the accuracy of detection of non-cavitated occlusal dentine caries among the diagnostic modalities used. Cone-beam computed tomography images were more sensitive, while digital radiographs were more specific.