X-ray AI Insights Navigator

An AI-powered web application for analyzing X-ray images using NVIDIA's advanced vision and medical language models. This tool provides educational insights and patient-friendly explanations of X-ray findings.

Important Disclaimer

This tool is for educational and informational purposes ONLY. It is NOT a substitute for professional medical advice, diagnosis, or treatment. AI interpretations can have errors or limitations. Always consult with a qualified healthcare professional for any medical concerns or interpretation of your X-rays.

Features

• AI Visual Analysis: Uses Meta's Llama 3.2 90B Vision model for detailed X-ray image analysis
• Patient-Friendly Explanations: Leverages Writer's Palmyra-Med-70B model for easy-to-understand medical explanations
• Medical Q&A Chat: Interactive chat feature for general medical questions
• PDF Report Generation: Download comprehensive reports with analysis and recommendations
• Specialist Finder: Integrated Google search to find relevant medical specialists
• Multi-format Support: Supports JPG, JPEG, and PNG image formats

Prerequisites

• Python 3.7 or higher
• NVIDIA API key (required for AI model access)

Installation

1. Clone or download the repository

   git clone <repository-url>
   cd Medical_Vision_X-ray

2. Install required dependencies

   pip install -r requirements.txt

3. Get NVIDIA API Key

   • Visit NVIDIA's API documentation
   • Sign up and obtain your API key

Usage

1. Start the application

   streamlit run Xray_Medical.py

2. Access the web interface

   • Open your browser and navigate to http://localhost:8501

3. Use the application

   • Enter your NVIDIA API key in the sidebar
   • Upload an X-ray image (JPG, JPEG, or PNG format)
   • Click "Analyze X-ray (Vision AI)" for initial analysis
   • Click "Get Patient-Friendly Explanation" for simplified interpretation
   • Download PDF reports as needed
   • Use the chat feature for general medical questions

File Structure

vlm_demo/
     Xray_Medical.py          # Main application file
     requirements.txt         # Python dependencies
     README.md               # This file
     logo.png                # Optional logo file
     sample_xray.jpg         # Optional sample X-ray image

Dependencies

The application requires the following Python packages:

• streamlit: Web application framework
• Pillow: Image processing library
• openai: OpenAI API client (used for NVIDIA API calls)
• fpdf2: PDF generation library
• requests: HTTP library for API calls

Configuration

Constants (configurable in Xray_Medical.py)

• MAX_ENCODED_IMAGE_SIZE_BYTES: Maximum image size (770KB)
• APP_NAME: Application title
• APP_LOGO_PATH: Path to logo file (optional)
• SAMPLE_IMAGE_FILENAME: Sample X-ray image filename

API Models Used

• Vision Analysis: meta/llama-3.2-90b-vision-instruct
• Medical Explanations: writer/palmyra-med-70b

Key Functions

Image Processing

• encode_image_to_base64: Converts images to base64 for API transmission
• format_vlm_output_for_display: Formats AI analysis output with emojis

PDF Generation

• generate_xray_report_pdf: Creates comprehensive PDF reports with analysis

API Calls

• call_llama_3_2_90b_vision_cached: Calls vision model for X-ray analysis
• call_palmyra_med_70b_cached: Calls medical model for explanations
• call_palmyra_chat_cached: Handles general medical chat

Security Notes

• API keys are not stored permanently
• All processing happens locally except for API calls to NVIDIA
• No medical data is permanently stored

Troubleshooting

Common Issues

1. API Key Errors: Ensure your NVIDIA API key is valid and has access to the required models
2. Image Size Issues: Reduce image size if upload fails (max ~580KB original size)
3. PDF Generation Errors: Ensure proper image format (JPG, PNG)

Error Handling

The application includes comprehensive error handling for:

• API connection issues
• Image processing errors
• PDF generation failures
• Invalid file formats

Contributing

This is an educational demonstration project. For improvements or bug fixes:

1. Ensure all changes maintain the educational/non-clinical nature
2. Test thoroughly with various X-ray image formats
3. Maintain proper error handling and user warnings

License

This project is for educational purposes only. Please respect NVIDIA's API terms of service and usage guidelines.

Support

For technical issues:

1. Check the console output for detailed error messages
2. Verify API key validity
3. Ensure all dependencies are properly installed
4. Check image format and size requirements

---

Remember: This tool is for educational demonstration only. Always consult qualified medical professionals for health-related decisions.