Hello Charlotte,
You can just use fslmaths to manually mask your input data by the 4D mask.
Kind Regards
Matthew
> Dear FSL experts,
>
> I want to run randomise to determine whether behavioural scores can explain a change in GM pre and post intervention, but using ROIs which are based on single subject anatomy. I have run longitudinal VBM analyses following the approach where each subject's scans are treated separately: VBM is run separately for time 1 and time 2 scans, fslmaths is then used to subtract one GM image from the other, resulting in a single volume which reflects the difference in GM between the two scan sessions - this image is what I'm looking at in randomise.
>
> I'd like to examine the correlation between a behavioural score and GM change, but I need to use a separate mask for each volume (i.e. to capture between subject anatomical differences in the ROIs). The masks have all been transformed into MNI space. I've tried making a 4D mask by concatenating all of the ROI masks but, having run a few sanity checks, I don't think that randomise is taking into account anything other than the first volume of the mask. Is there a way to do this please?
>
> Any advice would be great. Many thanks,
>
> Charlotte
|