A deep learning-powered web application for real-time dental disease prediction from X-ray images. This tool utilizes cutting-edge YOLOv8n/m and YOLOv10s object detection models to classify six dental conditions and assist clinicians in diagnosing oral health issues efficiently.
-
⚙️ YOLO-Based Detection: Utilizes Ultralytics' YOLOv8 and YOLOv10 models to detect and classify:
- Caries
- Infection
- Impaction
- Fractures
- Root canal issues
- Missing teeth
-
🎯 Model Performance:
- Achieved 92% mAP@0.5
- Recall: 0.8194
- Trained over 250 epochs on a custom-labeled dental X-ray dataset
-
🌐 Web Interface:
- Built using Vite + React for a fast and responsive front end
- Backend powered by Flask to serve predictions in real time
- Upload dental radiographs and receive immediate annotated results
| Layer | Technologies Used |
|---|---|
| Frontend | React (Vite), Tailwind CSS |
| Backend | Python, Flask |
| Model | YOLOv8n, YOLOv8m, YOLOv10s (Ultralytics) |
| Dataset | Custom-labeled dental X-ray images |