Comparison of fluorescence parameters between three generations of QLF devices for detecting enamel caries in vitro and on smooth surfaces

Abstract

BACKGROUND: This study compared two fluorescence parameters (fluorescence loss [ΔF] and red fluorescence gain [ΔR]) among three generations of quantitative light-induced fluorescence (QLF) systems with the aim of determining the validities of these parameters in the three devices for differentiating the severity of enamel caries.

METHODS: Forty-one extracted human premolars and molars with suspected enamel caries were selected. Fluorescence images of all teeth were obtained using first-, second-, and third-generation QLF systems (Inspektor Pro, QLF-D, and Qraycam, respectively). Fluorescence parameters were then calculated using proprietary software. All of the specimens were also categorized histologically using polarized-light microscopy (PLM) based on histological levels related to the lesion depth into sound enamel (S), caries limited to the outer half of the enamel (E1), and caries involving the inner half of the enamel (E2). The Mann-Whitney test with Bonferroni correction was used to compare fluorescence parameters among the three generations of systems. The sensitivity, specificity, and area under the receiver operating characteristics curve (AUC) at two thresholds (S/E1 for detecting enamel caries lesions and E1/E2 for differentiating the caries severity) were calculated for evaluating the validities of the fluorescence parameters obtained using all three generations of QLF devices.

RESULTS: ΔF did not differ significantly between the devices at any histological level. In addition, ΔF showed large AUCs at the thresholds of S/E1 and E1/E2 (0.97-0.98 and 0.89-0.90, respectively). On the other hand, ΔR was significantly higher for the third-generation device than for the first- and second-generation devices for E2 lesions (P < 0.001). At the S/E1 threshold, ΔR values of the first- and third-generation devices showed larger AUCs (0.96-0.97) compared with that of the second-generation device (0.91), whereas at the E1/E2 threshold the AUC was the largest for the third-generation device (0.87).

CONCLUSIONS: The ΔF fluorescence parameter did not differ between the three generations of QLF devices, and showed high validity values. In terms of ΔR, the devices of all generations also showed good diagnostic performance for quantifying and detecting enamel caries lesions, but the third-generation QLF system produced superior results.