kaurao / eigenmodes

This repository contains the code for reconstructing brain activity patterns as described in the paper Pang et al. 2023, Nature.

We also test parcel- and parcel+connectome-based basis vectors in addition to the eigenmodes provided by Pang et al. In addition, randomzied task data (spin permutation and Moran randomization) is also tested.

Requirements

• MATLAB 2021a or later (we use statistical abd parallel toolboxes)
• BrainSpace
  • git clone -b v0.1.10 https://github.com/MICA-MNI/BrainSpace.git
• matlab_GIfTI
  • git clone https://github.com/nno/matlab_GIfTI.git
• Rand_index
  • git clone https://github.com/cmccomb/rand_index.git
• Connectome Workbench
  • [matlab] curpath = getenv('PATH');
  • [matlab] setenv('PATH',sprintf('%s:%s','/path/to/.../workbench/bin_macosx64',curpath));
  • [matlab] clear curpath

Data

We use the data as provided by Pang et al.

• First clone the GitHub repo: https://github.com/NSBLab/BrainEigenmodes

• copy files from https://osf.io/xczmp/

  • To BrainEigenmodes/data/empirical

    • S255_tfMRI_ALLTASKS_raw_lh.mat
    • S255_high-resolution_group_average_connectome_cortex_nomedial-lh.mat

  • To BrainEigenmodes/data/results

    • basis_connectome_density_matched_midthickness-lh_evec_200.mat
    • basis_connectome_EDR_midthickness-lh_evec_200.mat
    • basis_connectome_midthickness-lh_evec_200.mat
    • basis_geometric_midthickness-lh_evec_200.txt

Usage

• [terminal] git clone https://github.com/NSBLab/BrainEigenmodes.git (add data as described above)
• [terminal] git clone -b v0.1.10 https://github.com/MICA-MNI/BrainSpace.git
• [terminal] git clone https://github.com/nno/matlab_GIfTI.git
• [terminal] git clone https://github.com/cmccomb/rand_index.git
• [terminal] git clone https://github.com/kaurao/eigenmodes.git
• [terminal] mkdir -p ./eigenmodes/results
• [terminal] cd ./eigenmodes
• [matlab] generate_figure;