connectome-analysis

Functional connectivity matrices are computed from ROI time series using correlation, with optional partial correlation, to produce FC matrices for downstream brain network analysis. The process uses nilearn and bctpy to transform fMRI signals into connectivity data.

Connectome analysis graph metrics include clustering coefficient, characteristic path length, global efficiency, modularity, and rich-club structure. These network science measures are calculated from thresholded adjacency matrices derived from functional connectivity data.

Hub detection in a brain network uses centrality-based heuristics to identify hub regions from functional connectivity matrices. The Skill computes these centrality measures from ROI time series to pinpoint highly connected nodes in the connectome.

Comparing functional connectivity between patients and controls is supported by computing group-level FC matrices and deriving network metrics for each group. The Skill processes ROI time series for both populations to enable statistical comparison of connectivity patterns.

Yes, bctpy and nilearn are required dependencies to compute rich-club organization and other graph metrics. The Skill also depends on networkx, numpy, pandas, scipy, and matplotlib to perform connectome analysis and generate visual outputs.

Evaluating resting-state functional connectivity involves computing FC matrices from ROI time series, thresholding them into adjacency matrices, and then calculating graph metrics like modularity and global efficiency. This Skill automates that pipeline using bctpy and nilearn.