Fine-grained spectral mapping of voxel-wise connectivity: A framework and application on resting-state brain fMRI data

Code accompanying the Ph.D. disseration of Serafeim LOUKAS entitled: "Multivariate and predictive models for brain connectivity with application to neurodevelopment".

URL: https://infoscience.epfl.ch/record/285683?ln=en, Thèse n. 8854 (EPFL, 2021)

---

main_modularity_acc.py: the main script for modularity estimation using Accordance as the weighted, undirected adjacency matrix of the underlying network

main_modularity_disc.py: the main script for modularity estimation using Discordance as the weighted, undirected adjacency matrix of the underlying network

main_laplacian_acc.py: the main script for normalized laplacian estimation using Accordance as the weighted, undirected adjacency matrix of the underlying network

main_laplacian_disc.py: the main script for normalized laplacian estimation using Discordance as the weighted, undirected adjacency matrix of the underlying network

Dependencies: numpy, scipy, os, nibabel and pandas

---

Important:

• To be able to run the example code you need to download the toy data from this link: https://drive.google.com/file/d/1Orq5wTBIIttZgo-_F9aG-YCbS4Hi4M0b/view?usp=sharing and put the .zip file into the Voxel_wise_Laplace_Modularity/Data/ directory that contains the GM mask. You should not unzip the data.

---

Note:

• For the core framework only numpy and scipy are needed.
• For loading some data and actually saving the eigenvectors as brain maps, os, nibabel and pandas are also needed.

---

Usage: In the main directory, launch python3 main_modularity_acc.py depending on the desired case.

---

The code for the overlay of the Yeo functional atlases onto the leading eigenmodes can be found here: https://github.com/seralouk/Brain_3D_Volume_Yeo_Mapping