🩺 IntelliMed — AI-Powered Longitudinal Health Intelligence...

🚀 Overview

IntelliMed is a premium React Native healthcare application designed to deliver AI-Powered Longitudinal Health Intelligence. It provides patients and clinical practitioners with highly accurate, structured clinical OCR interpretations, AI chatbot guidance, and longitudinal diagnostic history tracking.

Warning

Clinical Disclaimer: IntelliMed is designed strictly for informational and clinical tracking support purposes. It does not replace professional medical diagnosis, advice, or therapy. Always consult with a qualified primary care physician.

---

🏗️ System Architecture & Design

IntelliMed is engineered with a decoupled, highly scalable event-driven architecture designed to process medical reports asynchronously without blocking client APIs.

System Architecture Flowchart

graph TD
    A[React Native Expo App] -->|1. Upload File| B(FastAPI Gateway)
    B -->|2. Register Job| C[(SQLite/Postgres Database)]
    B -->|3. Push Task| D[Redis Message Broker]
    D -->|4. Pull Task| E[Celery Background Worker]
    E -->|5. Vision OCR Extraction| F[Google Gemini Vision API]
    E -->|6. Normalization & Sanity Validation| G[Master Biomarker Registry]
    E -->|7. Trend Analysis| C
    E -->|8. Generate Safe Interpretation| H[Gemini LLM]
    E -->|9. Finalize Job & Save| C
    A -->|10. Poll Status / Retrieve Results| B

Loading

🧱 Core Architecture Layers

1. Frontend Interface (React Native / Expo):
   • State-driven session validation utilizing secure AsyncStorage tokens.
   • Responsive layouts styled around the custom corporate palette (Navy, Teal, Amber Gold).
   • Conditional stack navigation preventing unauthorized routing transitions.
   • Dynamic offline simulation support for immediate testing and presentation.
2. API Gateway (FastAPI):
   • Asynchronous REST endpoints for ultra-low latency routing.
   • JWT credentials checking for user authentication.
   • Unified CORS handling for local Expo host routing.
3. Task Queue & Pipeline Broker (Celery + Redis):
   • Asynchronous file parsing pipeline to isolate resource-heavy OCR operations.
   • Redis serves as the rapid in-memory database broker.
   • Celery workers manage job status sequences (preprocessing, ocr_extracting, analyzing, completed, failed).
4. AI & Interpretation Pipeline (Gemini Pro & Vision):
   • Vision OCR: High-confidence text capture of scanned PDF/image reports.
   • Normalization Service: Standardizes clinical biomarker synonyms (e.g., matching "Glucose", "Fasting Sugar", and "FPG" to a single master record).
   • Clinical Boundary Safeguards: Rejects files with confidence scores < 85% or inputs that fail biological boundary ranges.
5. Database Storage:
   • SQL database (health_intelligence.db used for local development) managing tables for Users, Report, BiomarkerTrend, ProcessingJob, and PipelineAuditLog.

---

📸 Application Screenshot Gallery

Here are the actual visual previews of IntelliMed in action, showcasing the premium HSL Corporate Navy and Teal styling overhaul:

Screen Name	Preview Image	Key Functionality
Login Screen		Branded with the official logo, featuring offline bypasses for demo accounts.
Patient Home Dashboard		Welcome greeting, quick action triggers, and recent laboratory report history feed.
Analyzed Report Results		Asynchronous AI clinical interpretation card presenting diagnosis and suggested therapies.
Lifestyle Recommendations		Actionable lifestyle adjustments and longitudinal clinical advice parameters.
AI Clinical Assistant		Multilingual interactive guidance powered by custom clinical guidelines.
Report-Grounded Chat		Directly query details, diets, and metabolic indexes based on active report context.
Shareable Interpretation		Export, share results, and track detailed metadata indices.

---

📦 Installation & Setup

Frontend Installation

# Clone the repository
git clone https://github.com/audhee/IntelliMed.git
cd IntelliMed

# Install node dependencies
npm install

# Start the Expo Dev Server (with cache clear to index new logo assets)
npx expo start --clear

Backend Installation & Startup

# Navigate to the backend folder
cd backend

# Activate your Python virtual environment (Windows)
.\venv\Scripts\Activate.ps1

# Install requirements
pip install -r requirements.txt

# Start the FastAPI server
uvicorn app.main:app --reload --host 0.0.0.0 --port 5000

Running the Celery Worker

Ensure your Redis server is running, then launch the background worker in a separate activated terminal:

celery -A app.worker.celery_app worker --loglevel=info

---

🔑 Environment Variables

Create a .env file in the backend directory matching the following structure:

APP_NAME="IntelliMed API"
APP_ENV="development"
DATABASE_URL="sqlite:///health_intelligence.db"
REDIS_URL="redis://localhost:6379/0"
GEMINI_API_KEY="your_google_gemini_api_key_here"
JWT_SECRET="your_secure_jwt_secret_token_here"