README.md

DICOM Radiology Image Interpreter

This application allows users to upload DICOM files from radiology scans (such as CT, MRI, X-Ray) and provides interpretation using simulated AI models.

Features

• Upload multiple DICOM files
• View medical images with metadata
• Get AI-simulated interpretation of images
• Demo mode when PyDICOM is not available

Setup for Hugging Face Spaces

1. Create a new Space with the following structure:

   • app.py - Main application file
   • requirements.txt - Python package dependencies
   • packages.txt - System packages needed
   • README.md - This documentation file

2. Use the "Gradio" Space SDK

3. Set resource requirements:

   • At least CPU basic setting
   • Recommended: CPU upgrade or GPU if available

Dependencies

The application requires the following Python packages:

• gradio==3.50.0
• numpy==1.24.3
• matplotlib==3.7.2
• pydicom==2.4.0
• pandas==2.0.3
• Pillow==10.0.0

And system packages:

• python3-dev
• build-essential
• gcc
• libxslt1-dev
• zlib1g-dev
• libjpeg-dev

How to Use

1. Upload one or more DICOM files using the file uploader
2. Click the "Analyze Images" button
3. View the results in the gallery below

Notes

• The app supports multiple DICOM files for cases like CT scans with multiple slices
• The AI interpretation in this demo is simulated; a real app would use trained models
• Uploaded medical data is processed temporarily and not stored permanently
• If PyDICOM is not available, the app runs in demo mode with sample data