Manuscript: https://www.biorxiv.org/content/10.1101/2021.03.30.437729v2
Steps for using the code (requires SPM mounted on path):
- Extract eigenvariates (from preprocessed fMRI data you have) for each brain-region (as in the example files ses-func01.mat and ses-func02.mat) and store it in a .mat file. You will need to run dcm pipeline on these files for each subject seperately across all sessions.
- Run import_ts.m
- Run specify_dcm.m
- Run estimate_dcm.m
- Using the esimated dcm, form a .mat file containing parents (as in example file parents.mat).
- Run sys_id_analysis.m (Note - the manuscript shows a certain procedure for system identification which was originally scripted in python but now has been adapted to a simpler code in this script. Be careful to select the N4weighing to be SSARX to take feedback into account to replicate manuscript's results).