Skip to content

An AI-powered medical assistant for elderly users to check drug interactions using RAG (Gemini + Neo4j)

Notifications You must be signed in to change notification settings

yanhui-ma-dev/MediGuard-AI-Assistant

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🛡️ MediGuard AI Assistant

Bridging the Healthcare Gap with Gemini 2.5 Flash & Neo4j Knowledge Graphs

Python Flask Gemini Neo4j

🌟 Our Philosophy: Beyond "Patient" Safety

We believe that health is a daily journey, not just a response to illness. Most people managing health needs aren't "patients"—they are proactive individuals looking for clarity.

  • Breaking the Information Barrier: We believe professional medical insights should be a public right, not a specialized privilege.
  • Simplification for Equity: By making medical reports easy to understand, we empower everyone—regardless of their educational background—to make informed decisions about their own bodies.
  • Human-Centric Tech: We combine the "logic" of Graph Databases with the "empathy" of LLMs to create a tool that speaks your language.

🗺️ Roadmap

  • Phase 1: Core Drug-Drug Interaction (DDI) engine with Neo4j.
  • Phase 2: Real-time medication scanning UI.
  • Phase 3: Medical Report Translation (Jargon-to-Human) — Leveraging Gemini 2.5 Flash to simplify complex medical terminology into actionable daily insights.

🚀 Key Capabilities

1. Intelligent Drug-Drug Interaction (DDI) Analysis

Leveraging the structured logic of Neo4j, MediGuard maps complex relationships between various medications. It identifies potential risks not just by name, but by analyzing active ingredients and metabolic pathways, ensuring that daily health management is backed by rigorous data.

2. Medical Report Translation (Upcoming)

Aiming to bridge the information gap where medical reports are filled with intimidating terminology. This upcoming feature will use Gemini 2.5 Flash to "translate" professional reports into clear, empathetic, and human-readable language.

  • Understandable: No more confusing acronyms or Latin terms.
  • Actionable: Clear insights that help you make informed daily health decisions.

📺 Live Demo

MediGuard Demo Video Click the image above to watch how MediGuard simplifies healthcare.


📂 Project Structure

MediGuard-AI-Assistant/
├── frontend/         # UI Components (HTML5, Tailwind CSS, JS)
├── backend/          # API Engine (Flask, Gemini 2.5, Py2neo)
├── database/         # Knowledge Graph Data (CSV files)
├── requirements.txt  # Python Dependencies
└── README.md         # Project Documentation

About

An AI-powered medical assistant for elderly users to check drug interactions using RAG (Gemini + Neo4j)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors