Skip to content

AUDIT

AUDIT: An open-source Python library for comprehensive evaluation of medical image segmentation models and MRI datasets analysis

Documentation: https://caumente.github.io/AUDIT/

Source Code: https://github.com/caumente/AUDIT/

Welcome to the official documentation for AUDIT (Analysis & evalUation Dashboard of artIficial inTelligence), a a tool designed to provide researchers and developers an interactive way to better analyze and explore MRI datasets and segmentation models. Given its functionalities to extract the most relevant features and metrics from your several data sources, it allows for uncovering biases both intra and inter-dataset as well as within the model predictions.

🚀 Key Features

  • Robust evaluation: Extracts region-specific features and calculates a wide range of performance metrics.
  • Interactive visualizations: Includes a dynamic Streamlit-based web app for intuitive data exploration.
  • Highly customizable: Easily extendable for additional features and metrics tailored to your needs.
  • Friendly integration: Supports plugins and external libraries for advanced analysis (e.g., ITK-SNAP, pymia).
  • Open Source: Fully available on GitHub with comprehensive tutorials and examples.

📚 What You'll Find Here

This documentation is structured to help you get the most out of AUDIT:

  • Getting Started: Learn how to install AUDIT and set up your first project.
  • API Reference: Detailed reference for all library classes and methods.
  • Analysis modes: Explore the dashboard included in the web app.
  • Tutorials: Hands-on examples demonstrating common use cases.
  • About: Check latest AUDIT release and license terms.

🌟 Quick Start

The best way to get familiar with AUDIT and explore all its capabilities is through our interactive DEMO. You can find it at: https://auditapp.streamlit.app/.

Users will find an online wep app with pre-configured data to explore features and compare the accuracy of a set of medical image segmentation models. However, you could use AUDIT easily in your own computer by following a few little steps.

Install AUDIT

Directly from PyPI:

pip install auditapp

Alternatively, clone the repository and install it locally:

git clone https://github.com/caumente/AUDIT.git
cd AUDIT
pip install -r requirements.txt

Launch AUDIT App

Start the interactive web app using our default configuration for data visualization and exploration

Run the following command if you installed AUDIT directly from PyPI: 

auditapp run-app

Or alternatively, if you cloned the repository and install it locally, run:

python src/audit/app/launcher.py

That's it! You're ready to explore our default data and evaluate AI segmentations models with AUDIT. Go to Getting Started to learn more about how to use AUDIT with your own datasets and configurations.

🤝 Contributing

Contributions are welcome! Please check the contributing guide for details on how to report issues, suggest new features, or contribute code.

Feel free to reach out at Contact Us.