BrainSpace

BrainSpace enables analysis of macroscale cortical gradients in neuroimaging and connectomics datasets to quantify topographic principles of brain structure and function.

Key Features:

• Identification of Gradients: Identifies cortical gradients as multivariate manifolds that capture macroscale organization of brain structure and function.
• Gradient Alignment: Aligns identified gradients across datasets or conditions to enable comparison.
• Visualization: Provides tools to visualize gradients for interpretation and comparison of spatial patterns.
• Controlled Association Studies: Performs association analyses between gradients and other brain-level features under controlled conditions.
• Adjustment for Spatial Autocorrelation: Implements null models that account for spatial autocorrelation when testing associations.
• Implementations: Available implementations in Python and Matlab.
• Validation and Consistency: Includes validation experiments demonstrating consistency for functional and microstructural gradients.

Scientific Applications:

• Structural and Functional Brain Organization: Quantifies macroscale topographic principles across species, developmental stages, and aging processes.
• Disease Perturbations: Investigates how diseases alter patterns of cortical gradients.

Methodology:

Uses multivariate machine learning approaches to identify manifolds described as cortical gradients, supports gradient alignment across datasets, conducts controlled association studies, and applies null models that account for spatial autocorrelation.