Validity of MicroCT for in vitro detection of proximal carious lesions in primary molars

Objective: This study aimed to validate the MicroCT for detection of proximal carious lesions in primary molars, using histology as the gold standard. Methods: Forty-eight proximal surfaces of primary molars were examined. Two calibrated examiners conducted the examinations independently. Proximal surfaces were visually scored, using ICDAS. Bitewing radiographs, Micro-CT and histological analyses used caries scores: 0 = sound; 1 = outer enamel; 2 = inner enamel; 3 = not spread dentine; 4 = outer dentine; 5 = inner dentine. Axial and sagital images were used for Micro-CT analysis, whilst for histology, tooth sections (400 mm) were analyzed stereomicroscopically (x15). Results: Inter-examiner agreement ranged from 0.87 to 0.93 kappa coefficient (k). Histological analysis revealed a frequency of sound tooth surfaces (18.8%) enamel carious lesions (E1) (48%) and dentine carious lesions (D1) (33.3%). MicroCT showed high correlation with histology (r(s)0.88). At both diagnostic thresholds (E1 and D1), sensitivity and accuracy were higher for MicroCT. Inter- device agreement between MicroCT and histology was k = 0.81. No difference was found between MicroCT and histology as gold standards for detecting carious lesions using ICDAS. Conclusion: MicroCT can be used as a gold standard for detecting carious lesions in proximal surfaces in primary molars. (C) 2011 Elsevier Ltd. All rights reserved.