2006 Volume 49 Issue 6 Pages 725-730
The importance of diagnosing caries during the early stage has been increasingly recognized along with the increasing prevalence of the concept of minimal intervention in dentistry. However, conventional diagnostic systems are not necessarily adequate for this purpose, and the diagnosis of caries needs to be improved. The air of this study was to elucidate the relation between the property values of images of pit-and-fissure discoloration and conventional caries diagnosis. The possibility of caries diagnosis using discoloration imaging was also investigated. Fifteen molars with pit-and-fissure discoloration were used in this experiment. The occlusal surface of each tooth was photographed with a digital camera and the degree of discoloration was examined. The fractal dimension, the area in proportion to the whole occlusal surface and the formfactor of the discoloration were calculated using image analysis software. Each tooth was also examined with a DIAGNOdent® system and was conventionally diagnosed using clinical procedured, according to the caries depth. The relationships between fractal dimension, area proportion, formfactor, DIAGNOdent value and the clinical diagnosis were analyzed statistically. The fractal dimension, the area proportion, and the DIAGNOdent value tended to increase with the depth of the caries: Spearman's correlation coefficients between the clinical diagnosis and fractal dimension, the area proportion, and the DIAGNOdent value were 0.867, 0.862 and 0.724, respectively (p<0.01 for all comparisons). Significant correlations were also indicated between the fractal dimension and the area proportion, the fractal dimension and the DIAGNOdent value, and the area proportion and the DIAGNOdent value; Pearson's correlation coefficients were 0.874, 0.771 and 0.714, respectively (p<0.01). On the other hand, the formfactor was found to be independent of the caries depth. These results indicated that image analysis of molar pit-and-fissure discoloration would be useful for caries diagnosis.
