Radio Vision
Medical imaging AI platform for chest and brain disease detection.
~92.9% multi-label accuracy on chest X-rays
Problem
Detecting chest and brain diseases from medical images.
What I built
- Chest and brain disease detection models (ResNet-50 / U-Net) trained on 100k NIH chest X-ray images.
- A Python / FastAPI backend.
- A Next.js interface with detection results, a 3D viewer and a forum for radiologists.
Stack
- Next.js
- FastAPI
- Python
- Computer Vision
Outcome
~92.9% multi-label accuracy on chest X-rays.