Clinical cross-sectional diagnostic accuracy study of DIAGNOdent Pen and QrayCam Pro quantitative fluorescence for occlusal caries detection and ICDAS II-derived lesion categorization in permanent posterior teeth.

内容摘要

Early detection of occlusal caries in posterior teeth is important for minimally invasive care. Visual systems such as the International Caries Detection and Assessment System (ICDAS II) standardize clinical scoring, but adjunctive fluorescence-based tools may improve objectivity. This clinical cross-sectional diagnostic accuracy study evaluated DIAGNOdent Pen and QrayCam Pro metrics (&#x394;Fmax and &#x394;Rmax) relative to ICDAS II-derived lesion categories in permanent posterior teeth. This clinical cross-sectional diagnostic accuracy study included 64 adults who contributed 145 unrestored posterior teeth; the tooth was the analytical unit because all measurements were site-specific. After prophylaxis and air-drying, calibrated examiners assigned ICDAS II scores, obtained DIAGNOdent Pen peak readings, and captured QrayCam Pro images at the same sites. QrayCam Pro analysis yielded maximum fluorescence loss (&#x394;Fmax) and maximum red fluorescence gain (&#x394;Rmax, %). ICDAS II scores were grouped as E0 (0), E1 (1), E2 (2-3), and D1 (4-5). Spearman correlation and ROC analyses were performed. Thirty sound surfaces and 115 ICDAS II-detected lesions were analysed. Median DIAGNOdent values increased from 10.5 in E0 to 74.0 in D1, &#x394;Fmax decreased from 0.0% to -&#x2009;60.0%, and &#x394;Rmax increased from 0.0% to 70.0%. ICDAS II correlated strongly with DIAGNOdent (&#x3c1;&#x2009;=&#x2009;0.91) and &#x394;Fmax (&#x3c1; = -0.80) and moderately with &#x394;Rmax (&#x3c1;&#x2009;=&#x2009;0.67) (all p&#x2009;<&#x2009;0.001). For E0 vs. (E1&#x2009;+&#x2009;E2+D1), AUCs were 0.94 for DIAGNOdent, 0.97 for &#x394;Fmax, and 0.74 for &#x394;Rmax. For E2 vs. D1, AUCs were 0.93, 0.85, and 0.80, respectively. In this exploratory tooth-level analysis, DIAGNOdent Pen and QrayCam Pro metrics showed descriptive agreement with ICDAS II-derived clinical categories. Because analyses were not adjusted for within-participant clustering and formal pairwise AUC comparisons were not performed, the observed AUC patterns should be interpreted cautiously. The devices appear to provide complementary adjunctive information, and proposed cut-offs require external validation before clinical use.