🧠 MedMinds.AI

Breaking Medical Language Barriers with Artificial Intelligence

---

🏥 Hackathon Submission Overview

Project Name: MedMinds.AI
Category / Theme: Healthcare | AI for Social Good
Development Environment: Kiro IDE
AI Model / API Used: Google Gemini API

---

🔗 Links

• Documentation: https://docs.google.com/document/d/191HCAubNria8Mhnp9kbQ3x7hUktxGytJVdJxhT9qWm4/edit?usp=sharing
• Deployed link: https://medminds.onrender.com/
• YouTube Link: https://youtu.be/qj5wFumZI6Q

---

🩺 Problem Statement

Medical reports, prescriptions, and diagnostic results are usually written in complex medical terminology that is difficult for common people to understand.

Challenges faced by patients:

• ❌ Medical jargon is not user-friendly
• ❌ No clear indication of whether a condition is normal or critical
• ❌ Language barriers for non-English speakers

These issues can lead to confusion, anxiety, delayed treatment, and poor health awareness.

---

💡 Solution Overview

MedMinds.AI is an AI-powered healthcare assistant designed to simplify medical information and make it understandable for everyone.

What MedMinds.AI does:

• Accepts medical data as text
• Supports PDF uploads of medical reports

Using Google Gemini API, MedMinds.AI:

• 🧠 Converts medical jargon into simple layman language
• 🚦 Classifies health parameters as Critical / High / Normal / Low
• 📊 Provides an overall health severity summary
• 🌍 Supports multi-language explanations

⚠️ Disclaimer: This application is for general awareness only and does not replace professional medical advice.

---

🚀 Features

• 📄 Upload medical reports (PDF / TEXT)
• 🔍 OCR-based text extraction (Tesseract + PyMuPDF)
• 🤖 AI-powered medical report explanation
• 🟢 Highlights normal values
• 🔴 Highlights abnormal values
• 🌐 REST API support

---

🛠️ Tech Stack

• Backend: Flask (Python)
• AI: Google Gemini API
• Frontend: HTML, CSS, JavaScript
• Others: Flask-CORS, Pillow

---

📁 Project Structure

MedMinds/
│── app.py
│── README.md
│── requirements.txt
│── .env
│── routes/
│   │── __init__.py
│   │── explain_report.py
│── services/
│   │── __init__.py
│   │── ocr_service.py
│   │── gemini_service.py
│── static/
│   │── css/
│   │   ├── chat.css
│   │   ├── index.css
│   │── js/
│   │   ├── chat.js
│   │   ├── index.js
│   │── images/
│   │   └── doctor-illustration.png
│── templates/
│   │── index.html
│   │── chat.html

---

⚙️ Installation & Setup

1️⃣ Clone the repository

git clone https://github.com/your-username/medminds-ai.git
cd medminds-ai

2️⃣ Create virtual environment

python -m venv venv
venv\Scripts\activate   # Windows
source venv/bin/activate # Linux/Mac

3️⃣ Install dependencies

pip install -r requirements.txt

4️⃣ Setup Environment Variables

Create a .env file:

GEMINI_API_KEY=your_api_key_here
PORT=5000

5️⃣ Run the application

python app.py

App will run at: http://localhost:5000

---

📤 API Endpoint

POST /explain-report

Request:

• Form-data

  • file: PDF

Response:

{
  "patient_name": "Ramesh Kumar",
  "summary": "Blood sugar levels are high",
  "tests": [
    {
      "name": "Post Prandial Blood Sugar",
      "value": "240 mg/dL",
      "status": "High"
    }
  ]
}

---

🔐 Rate Limit Note

If you face 429 Quota Exceeded errors:

• Check Gemini API usage
• Reduce request frequency
• Upgrade API plan if required

---

🎯 Use Cases

• Patients understanding lab reports
• Rural health workers (ASHA)
• Hackathons & health-tech demos
• Medical awareness platforms

---

📜 License

This project is for educational and hackathon use.

---

✨ Built with the goal of making medical reports simple for everyone.