Two Stage Machine Learning Framework to Identify Periodontitis and Dental Caries

Background: Detecting dental caries and periodontitis through X-ray images is challenging due to image variability, overlapping anatomical structures, and subtle early-stage symptoms. Early diagnosis is essential for preventing disease progression and maintaining oral health. Methods: This research proposes a two-stage machine learning system for automatic and precise dental disorder diagnosis. The X-ray images are preprocessed and segmented using the fuzzy C-means clustering algorithm (FCM) to isolate regions of interest (ROI). Textural features are extracted via the Grey Level Co-occurrence Matrix (GLCM). A Support Vector Machine (SVM) classifier is used to classify the images into healthy or unhealthy, and further categorize unhealthy images into periodontitis, dental caries, and other dental conditions. Results: The proposed model achieved diagnostic accuracies of 98.8%, 98.5%, and 98.2% for different classes, demonstrating its high performance. Concluding Remarks: The proposed framework proves to be a reliable tool for enhancing clinical diagnoses and automating dental disease detection.