Appropriate model & software for resting-state data in a 2group and 3session design

Dear community,

I am investigating 2 groups (control vs treatment, cross-sectional). All participants were scanned at baseline (B), directly after treatment (F1) and again 2 weeks later (F2), i.e., longitudinal.

I would now like to investigate voxel-wise differences in brain activation.
It is obviously interesting whether the groups show differecnes after treatment (F1) and whether some long-term effects persist (F2). Also it would be interesting whether there is some interaction of time and group.

Does anyone have suggestions how to handle this the best and simlpest way? Has someone recommendation regarding good scripts/software?

Thanks for your help!


If you are using resting state data, you can’t really model brain activity, per se. Usually with resting state data, you would typically look at functional connectivity. If you have fMRIPrep derivatives, I would recommend using either the CONN toolbox (MATLAB GUI) or XCP_D (Python/Command Line).


1 Like

Hi Steven,

thanks for your reply.
Actually, I have already computed voxel-wise eigenvector centrality values. So what I really want to compare are whole-brain eigenvector centrality maps. Sorry, my bad, I thought it would be easier if I’d don’t go into the specifics.

I am not familiar with python so much unfortunately. But do you think the CONN toolbox can handle such a model with 2 groups of participants that were scanned at 3 time points?

Thanks again and have a great weekend!

Oh, if you already have the maps, you can just run a second level model in SPM (MATLAB) or Nilearn (python). I personally prefer Nilearn, but both can handle any analysis that can be specified by a general linear model, which should suit your needs.