Optimized persistence homological scaffolds of hemodynamic networks covary with MEG theta-alpha aperiodic dynamics

1. Install requirements

pip install -r requirements.txt

2. Repo structure

• data: Contains numpy matrices used in the manuscript.
  • regions_info.npy: a (3, 360) array containing the names (row 0), functional networks (row 1), and myelin content (row 2) of 360 Glasser regions.
• src: Source code.
  • fmri_pipeline.py: functions to compute persistence homological scaffolds, persistence centrality vectors, and degree centrality vectors. Assume functional connectivity matrices are already computed.
  • meg_pipeline.py: function to perform source localization and IRASA decomposition on MEG data. Assume all MEG data has been downloaded from the HCP database.
  • utils.py: helper functions.
• results: Contains Jupyter notebooks (.ipynb files) documenting the experimental results and analysis.