Classification of Dental Caries Level Using Conjugate Gradient Backpropagation Models

• Published in: 2023 International Seminar on Application for Technology of Information and Communication (iSemantic) pp. 204 - 208
• Main Authors: Jusman, Yessi, Nur'Aini, Masayu Alya, Puspita, Sartika
• Format: Conference Proceeding
• Language: English
• Published: IEEE 16.09.2023
• Subjects: Backpropagation; Dental Caries; Dentistry; Feature extraction; Gabor Filter; Gabor filters; Machine Learning; Multi-Layer Perceptron; Seminars; Teeth; Training
• DOI: 10.1109/iSemantic59612.2023.10295351

Oral diseases are among the most prevalent noncommunicable diseases, affecting approximately 3.5 billion individuals globally. Untreated caries of primary and permanent teeth, severe periodontal disease, edentulism (total tooth loss), and lip and oral cavity cancer are the most burdensome oral diseases and conditions. Untreated dental caries is the most prevalent noncommunicable disease and a significant global public health concern, accounting for approximately 2 billion cases of all oral diseases. Several studies have shown that machine learning can help read patient conditions more quickly and accurately. This research designs a classification system using Multi-Layer Perceptron to determine the level of dental caries. The dental image consists of four dental caries levels, then extracted using a Gabor filter. Multi-Layer Perceptron performance in classifying dental caries levels shows promising results. Conjugate Gradient Backpropagation with Powell-Beale Restarts and Conjugate Gradient Backpropagation with Fletcher-Reeves Updates were used for the classification stage. The training result shows that Conjugate Gradient Backpropagation with Fletcher-Reeves Updates has a better accuracy average than Conjugate Gradient Backpropagation with Powell-Beale Restarts. The highest accuracy achieved by Conjugate Gradient Backpropagation with Powell-Beale Restarts is 96,5%.