neurocluster

GitHub

Spatially constrained clustering for neuroimaging

Installpak::pak('bbuchsbaum/neurocluster')

Overview

neurocluster implements spatially constrained clustering methods tailored for neuroimaging data, enabling data-driven brain parcellation that respects spatial contiguity. It supports both hierarchical and iterative clustering approaches with connectivity constraints, and provides tools for assessing cluster stability across bootstrap samples or parameter settings. Multi-scale clustering allows parcellations to be explored at different levels of granularity.

Key Features

Spatially constrained hierarchical clustering with connectivity enforcement
Connectivity-based parcellation for data-driven brain mapping
Cluster stability assessment via bootstrap and resampling methods
Multi-scale clustering for exploring granularity hierarchies
Visualization of cluster boundaries on brain volumes or surfaces