A web-based board review platform for radiology residency programs, focused on genitourinary imaging. Designed for professor-led live teaching sessions with real-time resident participation.

• Session Builder — configure sessions by topic, modality, difficulty, image mix, and timer
• Live Presentation — professor screen with image viewer, reveal answer, teaching points
• Resident Participation — residents join via session code or QR code on any device
• Real-time aggregate — live bar chart of resident answers updates as they submit
• Session Summary — overall score and per-topic breakdown at end of session
• Radiopaedia Integration — import GU imaging cases as draft MCQs for professor review
• GU Taxonomy — 9 topic areas, 40+ subtopics covering the full ABR Core GU domain

• Python 3.12+
• Node.js 20+

bash run.sh

This starts:

• Backend API at http://localhost:8000 (auto-seeds 25 GU board questions)
• Frontend at http://localhost:5173

Professor dashboard: http://localhost:5173/dashboard Resident join page: http://localhost:5173/join API docs: http://localhost:8000/docs

cd backend
cp .env.example .env
# Edit .env to set your PROFESSOR_PASSWORD

1. Go to Dashboard → New Session
2. Select topics, number of questions, image mix, and optional timer
3. Click Build Session — you land on the Presentation page
4. Display the QR code on the projector — residents scan and join
5. Click Start Session
6. Use → (or arrow key) to advance, Space to reveal the answer
7. Teaching point and live resident answer distribution appear after reveal
8. At the end: End Session → View Summary

Create a .json file following the template in backend/data/seed_questions.json and upload via:

• Dashboard → Content → Import JSON, or
• POST /api/content/import-json (multipart file upload)

Configure API credentials in .env:

RADIOPAEDIA_CLIENT_ID=your_client_id
RADIOPAEDIA_CLIENT_SECRET=your_client_secret

Then call:

POST /api/content/sync-radiopaedia?topic=genitourinary&limit=20

Imported cases appear in Dashboard → Draft Questions for professor review. You must fill in the MCQ options and correct answer before activating.

1. Push to GitHub (already done)
2. Connect repo to vercel.com
3. Set root directory to frontend
4. Add environment variables:

   VITE_API_URL=https://your-backend.up.railway.app
   VITE_WS_URL=wss://your-backend.up.railway.app
   

1. Go to railway.app → New Project → Deploy from GitHub
2. Select board-review-GU repo, set root directory to backend
3. Add a PostgreSQL plugin from Railway dashboard
4. Set environment variables:

   DATABASE_URL=postgresql+asyncpg://...  (auto-set by Railway plugin)
   SECRET_KEY=<random 32-byte hex>
   PROFESSOR_USERNAME=admin
   PROFESSOR_PASSWORD=<your password>
   CORS_ORIGINS=["https://your-app.vercel.app"]
   

5. Railway uses Procfile for the start command automatically

Note: The codebase uses SQLite locally and PostgreSQL on Railway. Only DATABASE_URL changes — all SQLAlchemy models are database-agnostic.

docker-compose up --build

board-review-gu/
├── backend/
│   ├── main.py              # FastAPI app + WS endpoint
│   ├── models/              # SQLAlchemy ORM models
│   ├── schemas/             # Pydantic request/response schemas
│   ├── routers/             # auth, questions, sessions, answers, content
│   ├── services/            # session_builder, websocket_manager, radiopaedia
│   ├── ws/                  # WebSocket session handler
│   └── data/                # seed_questions.json + gu_taxonomy.json
└── frontend/
    └── src/
        ├── pages/           # Dashboard, Builder, Presentation, Summary, Resident, etc.
        ├── components/      # ImageViewer, AggregateBar
        ├── hooks/           # useSessionSocket, useSessionTimer
        ├── store/           # Zustand session state
        └── api/             # Axios clients