Dear FSL team,
we are running an ICA analysis on multisite VBM data and found strong site and age effects. So, the idea was to regress out these effects first (in a higher-level analysis), saving the residuals and then do the ICA on the residuals. Looking at previous posts a way to do this would be using fsl_glm. However, I am a bit confused about what input is needed there. Is fsl_glm able to read in multi-subject data and do a higher-level analysis (as feat is doing using the .fsf file) and if so, how? Or alternatively, is feat able to save the residuals?
Thanks for your help,