Stavroula Michou1,2, Ana Raquel Benetti3, Christoph Vannahme4, Pétur Gordon Hermannsson4, Azam Bakhshandeh3, Kim Rud Ekstrand3. 1. Department of Odontology, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark, stmi@sund.ku.dk. 2. 3Shape TRIOS A/S, Copenhagen, Denmark, stmi@sund.ku.dk. 3. Department of Odontology, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark. 4. 3Shape TRIOS A/S, Copenhagen, Denmark.
Abstract
OBJECTIVES: To develop an automated fluorescence-based caries scoring system for an intraoral scanner and totest the performance of the system compared to state-of-the-art methods. METHODS: Seventy-three permanent posterior teeth were scanned with a three-dimensional (3D) intraoral scanner prototype which emitted light at 415 nm. An overlay representing the fluorescence signal from the tissue was mapped onto 3D models of the teeth. Multiple examination sites (n = 139) on the occlusal surfaces were chosen, and their red and green fluorescence signal components were extracted. These components were used to calculate 4 mathematical functions upon which a caries scoring system for the scanner prototype could be based. Visual-tactile (International Caries Detection and Assessment System, ICDAS), radiographic (ICDAS), and histological assessments were conducted on the same examination sites. RESULTS: Most index tests showed significant correlation with histology. The strongest correlation was observed for the visual-tactile examination (rs = 0.80) followed by the scanner supported by the caries classification function that quantifies the overall fluorescence compared to sound surfaces (rs = 0.78). Additionally, this function resulted in the highest intra-examiner reliability (κ = 0.964), and the highest sum of sensitivity (SE) and specificity (SP) (sum SE-SP: 1.60-1.84) at the 2 histological levels where the comparison with visual-tactile assessment was possible (κ = 0.886, sum SE-SP = 1.57-1.81) and at the 3 out of 4 histological levels where the comparison with radiographic assessment was possible (κ = 0.911, sum SE-SP = 1.37-1.78); the only exception was for the lesions in the outer third of dentin, where the radiographic assessment showed the highest sum SE-SP (1.78). CONCLUSION: A fluorescence-based caries scoring system was developed for the intraoral scanner showing promising performance compared to state-of-the-art caries detection methods. The intraoral scanner accompanied by an automated caries scoring system may improve objective caries detection and increase the efficiency and effectiveness of oral examinations. Furthermore, this device has the potential to support reliable monitoring of early caries lesions.
OBJECTIVES: To develop an automated fluorescence-based caries scoring system for an intraoral scanner and totest the performance of the system compared to state-of-the-art methods. METHODS: Seventy-three permanent posterior teeth were scanned with a three-dimensional (3D) intraoral scanner prototype which emitted light at 415 nm. An overlay representing the fluorescence signal from the tissue was mapped onto 3D models of the teeth. Multiple examination sites (n = 139) on the occlusal surfaces were chosen, and their red and green fluorescence signal components were extracted. These components were used to calculate 4 mathematical functions upon which a caries scoring system for the scanner prototype could be based. Visual-tactile (International Caries Detection and Assessment System, ICDAS), radiographic (ICDAS), and histological assessments were conducted on the same examination sites. RESULTS: Most index tests showed significant correlation with histology. The strongest correlation was observed for the visual-tactile examination (rs = 0.80) followed by the scanner supported by the caries classification function that quantifies the overall fluorescence compared to sound surfaces (rs = 0.78). Additionally, this function resulted in the highest intra-examiner reliability (κ = 0.964), and the highest sum of sensitivity (SE) and specificity (SP) (sum SE-SP: 1.60-1.84) at the 2 histological levels where the comparison with visual-tactile assessment was possible (κ = 0.886, sum SE-SP = 1.57-1.81) and at the 3 out of 4 histological levels where the comparison with radiographic assessment was possible (κ = 0.911, sum SE-SP = 1.37-1.78); the only exception was for the lesions in the outer third of dentin, where the radiographic assessment showed the highest sum SE-SP (1.78). CONCLUSION: A fluorescence-based caries scoring system was developed for the intraoral scanner showing promising performance compared to state-of-the-art caries detection methods. The intraoral scanner accompanied by an automated caries scoring system may improve objective caries detection and increase the efficiency and effectiveness of oral examinations. Furthermore, this device has the potential to support reliable monitoring of early caries lesions.