Characterization of regional differences in resting-state fMRI with a data-driven network model of brain dynamics

This repository contains the code used in [1].

Structure

• Folder hcp-proc/ contains the code needed for processing of Human Connectome Project data (in order to get the structural connectome and resting state fMRI, both in Desikan-Killiany parcellation).
• Folder ndsvae/ contains the code for the inference method.
• Folder study/ contains the scripts and notebooks for the simulations, inferences, and postprocessing. Entry point is the Snakefile.

Environment

Python 3.8 with standard scientific and neuroscientific libraries is necessary. Use the environment file study/env.yml to prepare the conda environment.

License

This work is licensed under MIT license. See LICENSE.txt for the full text.

References

[1] Viktor Sip, Meysam Hashemi, Timo Dickscheid, Katrin Amunts, Spase Petkoski, Viktor Jirsa. Characterization of regional differences in resting-state fMRI with a data-driven network model of brain dynamics. Science Advances 9, eabq7547(2023). doi:10.1126/sciadv.abq7547