RESUMEN
The aim of this study was to evaluate the accuracy of cone beam computed tomography (CBCT) in the detection of noncavitated occlusal caries lesions and to compare this accuracy with that observed with conventional radiographs. 135 human teeth, 67 premolars and 68 molars with macroscopically intact occlusal surfaces, were examined by two independent observers using the CBCT system: NewTom 3G (Quantitative Radiology) and intraoral conventional film (Kodak Insight). The true lesion diagnosis was established by histological examination. The detection methods were compared by means of sensitivity, specificity, predictive values and accuracy. To assess intra- and interobserver agreement, weighted kappa coefficients were computed. Analyses were performed separately for caries reaching into dentin and for all noncavitated lesions. For the detection of occlusal lesions extending into dentin, sensitivity values were lower for film (0.45) when compared with CBCT (0.51), but the differences were not statistically significant (p > 0.19). For all occlusal lesions sensitivity values were 0.32 and 0.22, respectively, for CBCT and film. The specificity scores were high for both modalities. Interobserver agreement amounted to 0.93 for the CBCT system and to 0.87 for film. It was concluded that the use of the 9-inch field of view NewTom CBCT unit for the diagnosis of noncavitated occlusal caries cannot be recommended.