How kidkrishkode/fundusguardai works
FundusGuard AI is presented as a clinician-facing glaucoma screening and advisory platform. Its documented workflow is clinician onboarding, fundus-image submission, AI risk prediction and visual explanation, report generation, longitudinal tracking, and export. The sampled implementation chiefly establishes a Node/Express web application, server-rendered pages, security helpers, and Google Sheets-backed memory; no sampled source demonstrates the clinical image-analysis workflow, so those capability links are documented but unverified.
Main entry points: index.js Express server · memory.ts Google Sheets memory · views Advisor report · App.js Browser interface
Components 18
Web application
- Express server
index.js - Server-rendered pages
views - Shared templates
components - Browser interface
public/App.js
Clinical workflow
- User onboarding
views - Fundus submission
views - Glaucoma prediction
- XAI visualization
- Advisor report
views - Patient tracking
views - Report export
views
Platform services
- Security helpers
config/security.ts - Web helpers
config/hex.ts - Runtime settings
config/env-variables.ts
Persistence
- Google Sheets memory
config/memory.ts
Other components
- Clinician
- Google Sheets
- Python ML engine
Connections 20
- Clinician → Express server: uses
README.md - Express server → Server-rendered pages: renders
index.js - Express server → Shared templates: renders
index.js - Clinician → User onboarding: logs in
README.md - User onboarding → Fundus submission: uploads scans
README.md - Fundus submission → Glaucoma prediction: submits images
README.md - Glaucoma prediction → Python ML engine: uses models
README.md - Glaucoma prediction → XAI visualization: produces evidence
README.md - XAI visualization → Advisor report: informs report
README.md - Advisor report → Patient tracking: supports monitoring
README.md - Advisor report → Report export: exports summary
README.md - Browser interface → Express server: requests content
public/App.js - Browser interface → Express server: fetches legal content
public/App.js - Express server → Security helpers: decodes queries
index.js - Express server → Web helpers: renders templates
index.js - Express server → Runtime settings: loads settings
index.js - Express server → Google Sheets memory: uses shared state
index.js - Google Sheets memory → Security helpers: decodes credentials
config/memory.ts - Google Sheets memory → Web helpers: uses helpers
config/memory.ts - Google Sheets memory → Google Sheets: reads and writes
config/memory.ts