Generation of Teeth Caries Features for Human Dental Caries Classification

Abstract

Many dental diseases are experienced by humans, one of which is dental caries, there are three types of human dental caries, namely enamel caries, dentin caries and pulp caries. This study contains the detection of caries disease in human teeth using two-dimensional images and radiological results of x-ray periapical radiographs from a test image dataset that has a number of pixels between 374x288 to 672x514 pixels with an image resolution of 96 DPI. The original data of existing dental images was processed using Matlab language to obtain caries features through three stages of the processes: pre-processing stage which are stages of the preprocessing process that converts data from a two-dimensional color image (row/height, column/width) that is stored using three channels Red, Green and Blue (RGB), into a grayscale image with one channel, the process of extracting dental caries features by performing calculations caries area and calculate the distance of the caries area to the nerve canal (pulp), and the process of building learning or reference data from dental caries using 24 radiograph periapical data on molar tooth images processed using Matlab. Dental caries features extraction process and the features learning process to generate references features from dental caries is the main objective of this research. This study result was references features for human dental caries classification.