A Multi-Model Approach to Oral Pathology Classification and Dental X-Ray Analysis

The timely detection and accurate localization of oral diseases, such as oral cavity cancer and dental anomalies, is critical for improving preventive dental care and patient outcomes. This paper presents an AI-based system that employs a dual-architecture approach for analyzing oral and dental x-ray images. An ensemble of ResNet-50, EfficientNet-B1, and InceptionResN etV2 is used for oral disease classification, leveraging their complementary strengths to ensure accurate predictions across various conditions. For dental X-ray analysis, the Detection Transformer (DETR) model is utilized to detect and localize abnormalities in complex radiographic structures. This integrated system enhances diagnostic accuracy and provides early detection, treatment suggestions, and clinical decision support through a user-friendly interface and chatbot. The proposed approach improves accessibility and efficiency in oral health diagnosis, contributing to better patient outcomes and facilitating early intervention in dental care.