Hello,
One thing that may help is to reduce the brain-background % value to 0. You can also add a fixed term to data, this should be fine ( although if the data is strongly negative, you still want to set the brain/background % to 0 since the "background will now have higher intensities than some brain voxels )
Kind Regards
Matthew
> I'm trying to run a ROI to whole-brain connectivity analysis on resting state data using FEAT. I created the ROI time-course file using fslmaths, and have put this into my model as 1 EV. However, as I'm trying to run the analysis I get an error message that FEAT cannot create a mask image on my pre-processed data, presumably because there are negative values in the data (described in the post linked below).
>
> I looked at my data at the step just prior to removing covariates (so they have been slice-time-corrected and realigned), and there are a few negative values in this data, so I don't think it's only coming from regressing out the covariates. But there are many more negative values in my data that are fully pre-processed (with covariates removed) than the raw 4D files. Does anyone have suggestions on how to proceed so that I could run the analysis in FEAT? I'm willing to re-pre-process my data, but I want to make sure it will work before devoting the time and server space. Thanks!
>
> Catherine Burrows
>
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1501&L=fsl&P=R92847&1=fsl&9=A&J=on&d=No+Match%3BMatch%3BMatches&z=4
|