Lacuna

A Python package for advanced brain lesion analysis.

Lacuna bridges the gap between individual lesion masks and normative brain data. It provides a reproducible workflow for lesion network mapping and regional damage quantification, using BIDS-style naming conventions for input and output organization.

This project is under active development and has not yet been fully validated. Use with caution.

Analyses

• Functional lesion network mapping

  ---

  Map the functional brain circuitry linked to a lesion using resting-state functional connectivity.

• Structural lesion network mapping

  ---

  Map the structural disconnectivity of a lesion using normative tractogram data.

• Regional damage

  ---

  Quantify regional damage by measuring lesion overlap with standard brain parcellation atlases.

Install

pip install git+https://github.com/m-petersen/lacuna

Usage

# Create a tutorial dataset with synthetic lesion masks
lacuna tutorial my_dataset

# Fetch the HCP1065 tractogram
lacuna fetch hcp1065 --output-dir connectomes

# Run structural network mapping
lacuna run snm my_dataset output \
    --connectome-path connectomes/hcp1065.tck \
    --mask-space MNI152NLin6Asym

For the full walkthrough, see the Getting started tutorial. Note that Lacuna expects lesion masks to be in MNI space.

Documentation

• Tutorials

  ---

  Step-by-step Jupyter notebook tutorials covering each analysis type.

• How-to guides

  ---

  Practical guides for specific tasks beyond core analysis workflows.

• Reference

  ---

  CLI commands, options, and auto-generated API documentation.

• Explanation

  ---

  Background knowledge for using the package.

Issues

Please report issues on GitHub.

Meta VCI Map Consortium

This toolbox is developed as part of ongoing efforts within the Meta VCI Map Consortium, an international collaborative platform dedicated to advancing multicenter lesion analysis in vascular cognitive impairment. The consortium brings together large-scale datasets and interdisciplinary expertise to improve reproducibility and generalizability in lesion–symptom mapping and related approaches. This toolbox reflects these principles by providing standardized, scalable tools for the analysis of lesion–behavior relationships across diverse cohorts.

Funding