How buu1210/mammo-ai works
MAMMO-AI is a planned web application for preliminary mammogram analysis and guided practice; the repository is currently in planning and design, and the app, model, and tutor are marked as pending construction. The documented analysis flow runs from browser upload through an API to classification and Grad-CAM, with results persisted and returned to the app. A separate tutor workflow uses reference images and persists practice sessions. These are documented design relationships, not verified runtime calls; the API has no implementation path in the tree.
Main entry points: tutor AI tutor · app Web app · mammo_ai_bd.sql PostgreSQL database · modelo CNN classifier
Components 7
Image analysis
- Web app
app - Web API
- CNN classifier
modelo - Grad-CAM
modelo
Guided practice
- AI tutor
tutor
Data persistence
- PostgreSQL database
database/mammo_ai_bd.sql
Other components
- Health user
Connections 10
- Health user → Web app: uploads image
README.md - Web app → Web API: sends image
README.md - Web API → CNN classifier: requests classification
README.md - Web API → Grad-CAM: requests heatmap
README.md - CNN classifier → Grad-CAM: shares model
README.md - Web API → PostgreSQL database: stores analysis
README.md - Web API → Web app: returns result
README.md - Health user → AI tutor: practices scenarios
README.md - AI tutor → PostgreSQL database: reads references
README.md - AI tutor → PostgreSQL database: saves sessions
README.md