Objective: This study compared four caries detection methods to detect occlusal caries: the visual criteria "International Caries Detection Assessment System" (ICDAS), DIAGNOdent (DD; Kavo; Germany), Quantitative Light-induced fluorescence (QLF; Inspektor Research Systems, The Netherlands), and neksDuo (ND; NEKS Technologies Inc.; Canada). Methods: Seventy human posterior teeth were selected based on the ICDAS criteria, with 10 teeth on each of the 7 categories. The teeth were cleaned with a rotating brush, autoclaved and stored in water at -4°C. A digital image (1:1 magnification, FUJIFILM FinePixS2Pro) was obtained of each occlusal surface, a site was selected and marked on the image. All sites were examined using DD, QLF, and ND methods. Thereafter the teeth were hemi-sectioned and examined wet under a digital stereomicroscope (1x magnification, Nikon SMZ1500, NIKON INC.). The sections were classified into 5 categories according to Ekstrand et al criteria (1998). Results: Specificity was significantly higher for ND (92%) and DD (92%) than ICDAS (58%), QLF area (42%), QLF dF (17%), and QLF dQ (25%). Overall sensitivity scores were significantly higher for QLF area, dF and dQ (100%), and ICDAS (95%) than DD (84%) and ND (83%). Sensitivity was significantly higher for lesions in outer enamel for QLF area, dF and dQ (100%) than ICDAS (67%), DD (33%), and ND (22%); and ICDAS had significantly higher sensitivity than ND. There were no significant differences between any of the methods for any of the other histology categories. Area Under the ROC Curve (AUC) was higher for ICDAS (0.95) followed by QLF dF (0.93), DD (0.92), QLF dQ (0.91), QLF area (0.90), and ND (0.89).Conclusion: The best combinations of sensitivity and specificity were for ND and DD and the best AUC was for ICDAS.

Back to the AADR/Pfizer Hatton Awards Program
Back to the IADR/AADR/CADR 85th General Session and Exhibition (March 21-24, 2007)

Top Level Search