Logit-Based Artificial Bee Colony Optimization (LB-ABC) Approach for Dental Caries Classification Using a Back Propagation Neural Network

Abstract

Oral pain caused by bacterial infection, or caries, is an issue that can significantly impair individuals’ ability to function in their day-to-day lives. Analysis of dental caries using X-ray images is tricky, and dental professionals are struggling to find a better solution in order to avoid misclassification of dental caries stages and potential false diagnosis. To avoid such classification and diagnostic inaccuracy, a method is proposed in this work that utilizes a hybrid approach combining a logit-based artificial bee colony optimization algorithm [LB-ABC] with a back-propagation neural network. This approach is implemented to boost the back-propagation algorithm for a proper training and testing process, thereby attaining the highest classification accuracy by using dental X-ray images as the numerical input generated through a gray-level co-occurrence matrix (GLCM), a texture feature extraction process. With this approach, the proposed work achieved an optimal accuracy of 99.16% and a minimized error of about 0.0033.