Assessment of the correlation between fluorescence-featured intraoral scanner, laser fluorescence and spectrophotometric analyses in caries-affected dentin: an in-vitro diagnostic accuracy study

To investigate the relation between the caries scoring system of the fluorescence intraoral scanner (IOS) Trios 4 (3Shape, Denmark)a red laser light-induced fluorescence device DIAGNOdent Pen (Kavo, Germany), and the color parameters (L*a*b*) by a clinical contact type hybrid spectrophotometer Rayplicker (Borea, France) in the assessment of caries affected dentin tissue. Caries lesions were minimally invasively removal from 186 extracted molars. Teeth were scanned using Trios4, and the integrated software scored the cavity floor regarding the colors (yellow, initial caries; red, moderate-extensive caries) depending on the fluorescence features. Then the DIAGNOdent measurements were obtained from each cavity's deep and discolored surface points following the previously obtained Trios4 colors. Thus, the Trios4 reading could be quantitatively assessed. Cross-polarization photographs were using the Rayplickerand L*a*b color parameters, and 3D Master (VITA) color mapping was obtained by using the RayPlicker's software. Spearman's Rho Correlation, Kappa, Mann Whitney-U, One-way Analysis-of-Variance, and Kruskal Wallis tests were used for the statistical analyses (P < 0.05). Sensitivity, specificity and AUC were calculated. Trios4 and DIAGNOdent were positively correlated (r = 0.733;P < 0.001). The L*parameter by Rayplicker and Trios4 was negatively correlated(r=-0.742;P < 0,001). The a*parameter by Rayplicker and Trios4 scoring positively correlated(r = 0.552;P < 0.001). No significant correlation was observed between b*parameter by Rayplicker and Trios4 (r = 0.023; P = 0.760). DIAGNOdent readings according toTrios4 scorings were significant (P < 0.001). The teeth without caries scored by the Trios4 corresponded to the median value of 29 in DIAGNOdent readings, and 88 for the teeth with caries (by Trios4 / yellow-red). The agreement between the Trios4 and DIAGNOdent was 52.2% for the specimens with no residual caries. The presence of caries scores by the Trios 4 corresponded to the DIAGNOdent readings of %100 for all the teeth evaluated. Regarding the RayPlicker assessments, 33% of the sound cavities corresponded to 2M3 color, and 73% corresponded to 5M3 color. Trios4 scorings presented 0.782 AUC, 56.30% sensitivity, and 100% specificity in DIAGNOdent reference. Trios4 scoring was considered coherent with the DIAGNOdent Pen. DIAGNOdent readings and the Level of L* and a* parameters in the dentin tissue might be considered interrelated. Fluorescence-featured scanner devices can be useful clinical tools to evaluate remaining dentin tissue during the caries removal procedure.