CaliBrain documentation

CaliBrain is a Python library designed for simulating realistic Magnetoencephalography (MEG) and Electroencephalography (EEG) data and evaluating the calibration and uncertainty quantification of brain source estimation algorithms.

It provides tools to:

• Simulate source activity, leadfield matrices, and sensor-level measurements with controllable noise levels and source configurations (fixed or free orientation).

• Estimate and visualize confidence intervals for source estimates based on posterior covariance information.

• Analyze the calibration of source estimation methods by comparing confidence levels to the actual proportion of ground truth values captured.

This documentation provides installation instructions, usage examples, and a detailed API reference.

Documentation

• Installation
  • Pip Installation (Recommended)
  • Simple Pip Installation via requirements.txt
  • Conda Installation
  • Which method should I use?
  • Minimum Requirements
  • Optional Setup for Development
• Usage
  • Leadfield Simulation
  • Running Experiments
  • API Overview
  • Examples
  • Summary
• API Reference
  • Leadfield Simulation
  • Data Simulation
  • Source Estimation
  • Uncertainty Estimation
  • Utilities
  • Benchmarking
• Examples
  • Leadfield Simulation Example
  • Running Experiments Example
• Read Me
  • Overview
  • Calibration Tasks
  • Main Parameters
  • Outcomes
  • Installation
  • Usage
  • Contributing
  • License
  • Citation
• Results
  • 1. benchmark_results/
  • 2. figures/
  • 3. forward/
• General Notes
• Contributing
  • Getting Started
  • How to Contribute
  • Code Style and Quality
• Changelog
  • Version 1.0.1 (2025-05-24)
  • Version 0.1.0 (2025-04-28)