Maulisa Oktiana, Maya Fitria, Chesyiel Nashrifa, Khairun Saddami, Rizkika Putri, Rizki Novita
Dental caries is a progressive disease that attacks the hard tissues of teeth and can cause serious complications such as chronic pain, tooth loss, systemic infections, and periodontal disorders. Early detection is crucial to prevent further damage, but conventional examination methods, which still rely on manual observation, are subjective and potentially lead to misdiagnosis. One of the main challenges in automatic dental image classification is low image quality, especially in terms of contrast and clarity of tooth anatomical structures. Therefore, image quality enhancement techniques are a crucial component in the classification stage, as they can improve the visibility of important features and support the performance of deep learning models. This study developed a classification model for normal and caries-prone tooth images using the ResNet50 architecture, integrated with the Contrast Limited Adaptive Histogram Equalization (CLAHE) image enhancement technique. The CLAHE technique is used to enhance local contrast, making tooth structures clearer and more easily recognizable in the model. Test results show that the application of CLAHE increases classification accuracy by up to 84% at a learning rate of 10-4, compared to 79% without CLAHE. These findings confirm that image enhancement techniques play a crucial role in enhancing the accuracy, consistency, and reliability of automated classification systems for dental caries detection. ©2025 IEEE.
Dept. of Electr. and Comp. Eng., Univeristas Syiah Kuala, Banda Aceh, Indonesia; Dept. of Dental and Oral, Regional General Hospital dr. Zainoel Abidin (RSUDZA), Banda Aceh, Indonesia; Faculty of Dentistry, Universitas Syiah Kuala, Banda Aceh, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock