Medical AIComputer VisionReactFlask
OsteoAI
SOTA Osteoporosis Detection from X-Ray Images
Project Dossier
A full-stack medical application for osteoporosis detection from X-Ray images. The system uses fine-tuned image models to classify bone density health with high accuracy.
Engineering
Deep Learning Pipeline
Fine-tuned 3 distinct Deep Learning models for bone disease classification. Implemented a regression head for T-Score prediction, ensuring results fall within a ±0.25 unit margin of error 85% of the time.
Full-Stack Deployment
The model is served via a Flask API, integrated with a React frontend, and deployed on Azure Web Services for high availability.
Key Stats
94%
Accuracy
0.85
R2 Score
0.25
MAE
Actions