Perceived Diagnostic Value of Fluorescence-Enhanced 3D Imaging for Detecting Caries Adjacent to Restorations: A...

Background/Objectives: Caries adjacent to restorations remain a leading cause of restoration failure and replacement. Conventional diagnostic methods are limited by subjectivity and restricted visualization. Fluorescence-enhanced three-dimensional (3D) imaging has been proposed to improve detection accuracy, but evidence on its clinical perception and usability remains scarce. The objective of this study was to evaluate the perceived diagnostic value of fluorescence-enhanced 3D imaging in detecting caries adjacent to direct restorations. Methods: A cross-sectional questionnaire-based survey was distributed to undergraduate dental students and licensed dentists (n = 94). Participants assessed images of extracted teeth with direct restorations presented in three formats: conventional photographs, monochromatic 3D models, and 3D models with fluorescence. Responses were analyzed using descriptive statistics, chi-square tests, and Cohen's kappa to measure inter-rater agreement. Results: Overall, 64.9% of respondents reported that fluorescence-enhanced images improved their diagnostic decision-making, while 29.8% reported partial benefit. Fluorescence was mainly perceived as helpful in defining cavity margins (53.3%) and assessing lesion volume (42.4%). Most participants preferred 3D models with fluorescence over conventional images for diagnostic value. However, inter-rater agreement was generally poor (κ range: -0.05 to 0.25; median κ = 0.02; only 4 images showed weak but statistically significant agreement), with only a few images demonstrating weak but statistically significant agreement. Notably, 39.3% of participants reported prior experience with 3D imaging, which was associated with greater confidence in interpreting fluorescence-enhanced images. Participants with prior 3D imaging experience reported greater confidence in fluorescence interpretation. Conclusions: While fluorescence-enhanced 3D imaging is perceived as a useful adjunct for visualizing lesion margins and depth, it does not currently yield consistent diagnostic agreement across clinicians. Training, calibration, and integration of artificial intelligence support may enhance the clinical reliability of this technology.