help > A question about ABIDE ROI
Feb 10, 2022  01:02 PM | Matthew Ryan - The University of Adealide
A question about ABIDE ROI
Hey all,

I am looking at doing some research with the ABIDE dataset (specifically the NYU site) and I am having trouble getting my head around the way the ROI time series have been calculated.  I have attempted to calculate the ROI time series myself for the AAL atlas with python (using the nilearn package), MATLAB (using spm12 and the AAL toolbox) and with R (very hack and slash).  These all return somewhat similar results to each other.

My issue is that none of these methods return anything remotely similar to what is provided for the AAL ROI downloads from the ABIDE initiative.  For example,  using subject 50954 the time series I am able to extract for the Precentral_L region are all valued in the 60's, while the AAL dataset downloaded directly from ABIDE has values around 13.

I assume that I am missing some sort of confounds to regress out or some other silly mistake, but I can't get my head around it.  Any insight or help would be greatly appreciated.

Kind regards,