HealthAI Platform

A full-stack AI-powered healthcare platform for doctors and patients featuring symptom analysis, image diagnosis (pneumonia, skin cancer, anemia), OCR medical report storage, blockchain, and analytics dashboards with Supabase integration.

Project Structure

• Backend (Flask): app.py, run.py, services in services/, AI models in ai_models/, Supabase client in database/.
• Frontend (Next.js 13 app router): UI pages in app/, shared components in components/, utilities in lib/, API routes in app/api/.
• Config: config.py, environment variables, Tailwind/Next configs.

Prerequisites

• Python 3.10+
• Node.js 18+ and pnpm (or npm)
• Supabase project (URL + anon/service keys)
• Tesseract OCR installed locally (for OCR features) and OpenCV runtime deps
• Optional: Docker if running in a container

Environment Variables

Backend (.env or shell):

• SUPABASE_URL, SUPABASE_KEY
• PINATA_API_KEY, PINATA_SECRET_KEY (IPFS)
• OPENAI_API_KEY, HUGGINGFACE_API_KEY (AI features)
• SECRET_KEY, ENCRYPTION_KEY
• UPLOAD_FOLDER (defaults to uploads)
• FLASK_ENV=development, FLASK_DEBUG=1 (for local dev)

Frontend (.env.local):

• NEXT_PUBLIC_SUPABASE_URL
• NEXT_PUBLIC_SUPABASE_ANON_KEY

Installation

Backend

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Frontend

pnpm install
# or npm install

Running

Backend (Flask)

source .venv/bin/activate
python run.py
# Serves on http://localhost:3000 by default

Frontend (Next.js)

pnpm dev
# or npm run dev
# Serves on http://localhost:3000 (adjust if backend conflicts, e.g., set FRONTEND_PORT=3001)

If port conflict occurs (backend also defaults to 3000), run one of them on a different port:

# Backend custom port
FLASK_RUN_PORT=5000 python run.py
# Frontend custom port
PORT=3001 pnpm dev

Key Endpoints (Backend)

• POST /api/auth/wallet-login
• POST /api/ai/symptom-analysis (app.py)
• POST /api/ai/image-diagnosis (supports diagnosis_type like pneumonia, skin_cancer, eye_anemia)
• POST /api/ai/ocr-analysis (OCR + report analysis)
• POST /api/ai/anemia/eye, /api/ai/anemia/nail
• GET /api/ai/anemia/health
• POST /api/medical-records
• GET /api/analytics/dashboard

Frontend Highlights

• Patient dashboard: app/patient-dashboard/page.tsx
• Analytics: app/analytics/page.tsx
• AI diagnosis (anemia): app/ai-diagnosis/anemia-detection/page.tsx
• AI symptom checker: app/ai-symptom-checker/page.tsx
• Blockchain UI: app/blockchain/page.tsx
• Medical records UI/API: app/medical-records/add/page.tsx, app/api/medical-records/route.ts

Data & Storage

• Supabase is used for auth/data (services/analytics_service.py, utils/validators.py).
• File uploads stored under uploads/ (see app.py); IPFS planned via Pinata.

AI Models

• Disease predictor: ai_models/disease_predictor.py
• Image analyzer: ai_models/image_analyzer.py
• OCR processor: ai_models/ocr_processor.py
• Anemia detection pipeline: backend/anemia_detection.py

Styling & Config

• Tailwind: tailwind.config.ts
• Next config: next.config.mjs
• TypeScript config: tsconfig.json

Tips for Smooth Runs

• Ensure uploads/, models/, and logs/ directories exist (created automatically in run.py).
• Install Tesseract and OpenCV system dependencies if OCR/image analysis fails.
• Keep frontend and backend on separate ports to avoid conflicts.
• Set Supabase env vars for both backend and frontend before running.
• Use pnpm dev for faster frontend install/build; ignoreBuildErrors and ignoreDuringBuilds are enabled in next.config.mjs for local development.