Hej.
I have ICA results estimated using the GIFT toolbox that I am currently trying to load into a higher level FEAT analysis to test for a between-group difference in a covariate. I have extracted the spatial component of interest from the '4D' files (GIFT saves out the spatial maps into 4D files for each subject, where the fourth dimension corresponds to the number of estimated components) and saved these into individual subject folders (e.g., sub001.feat, sub002.feat, etc.).
I am able successfully set everything up in terms of selecting the data and specifying the model and contrasts, however, when I go to run the GLM, I keep getting an error saying that,
"Registration has not been run for all of the FEAT directories that you have selected for group analysis. Please turn on and set up registration."
The registration tab is grayed out for higher-level analyses, so I am unsure what to do at this point. The data has already been normalized to the MNI template using SPM12 prior to analysis in GIFT.
Any suggestions would be helpful, including if this is a fools errand and FEAT can only be used on FSL-created data.
Thanks in advance,
Suzanne
|