Image Processing Techniques for Caries Detection

Abstract

Dental caries or cavities are usually known as tooth decay, are affected by a breakdown of the tooth enamel. Although dental caries is generally preventable, they are found commonly in all age groups. If caries is untreated, can lead to a severe toothache, infection and tooth damage. Existing approaches using image processing are not too much efficient and leading dental experts to rely on conventional methods of visual assessments. In its place experts are still believing in an old-style visual or visual-tactile examination. newlineKeeping all these perspectives in mind, novel and unique caries detection technique using image processing is proposed. Initially, the dental images are pre-processed by means of popular techniques like contrast enhancement, grey thresholding and active contour for better visualizations of affected areas. Subsequently, feature extraction is carried out with the aid of dimensionality reduction technique which is built on the concept of Principal Component Analysis. It is a modified version of Principal Component Analysis termed as Multilinear PCA. newlineAlong with dimensionality reduction, Multilinear PCA is preserving original structural details of the image with covariance. Afterward, classification is achieved using Neural Network Classifier and Caries identification is accomplished. newlineIn the first contribution Multilinear PCA along with Neural Network (MPCA-NN), the main focus is to detect the caries on the basis of the advanced image processing algorithms in order to ensure detection accuracy. The proposed feature extraction algorithm describes the images, which are two dimensional, in multidimensional space. Thus, the contents of the images are projected into multi-dimensional format and so the useful information from the images can be required. Moreover, the Neural-Network classifier is trained using enhanced learning algorithm which will be used to classify grade of caries.