 |
 |
Yes - but you easily can create a mask with the voxels of no interest simply set to zero, e.g. by using fslmaths using the -thresh and -bin options. Cheers, Cornelius On Tue, Sep 4, 2012 at 2:14 PM, Maren Strenziok <[log in to unmask]> wrote: > I thought that -m is for inclusion masks. Maren > > > On Sun, Sep 2, 2012 at 7:59 AM, Cornelius Werner > <[log in to unmask]> wrote: >> >> Hi, >> >> you can provide randomise with a mask by using the -m switch. >> >> Cheers, >> Cornelius >> >> Am 01.09.2012 um 03:48 schrieb Maren Strenziok: >> >> > Hi, >> > >> > I have functional connectivity maps (zstats maps) from 2 groups >> > (trainiedn/non-trained) and 2 time points (pre-training/post-training) and >> > was not able to get significant results looking at interaction effects >> > between group and time. I try to explore less strict comparisons now. Can >> > someone tell me whether it is possible to assess differences between the 2 >> > groups post-training while controlling for pre-training (baseline) >> > differences? I tried to add a voxel-dependent EV to my 2 samples t-test >> > design but got an error. In think I know what to put in the randominse >> > command but the set up of the design.mat and design.con files failed. Also, >> > what exactely does this do? My goal is to restrict analyses to fewer voxels, >> > for example only those for which there were no baseline differences between >> > groups. Alternatively, I am thinking of using a white matter skeleton map >> > (from tbss) to restrict data analysis only to voxels that are not in the >> > skeleton. I guess my main quesion here is, how do I set up an exclusion mask >> > in randomise? >> > >> > Maren >> > >> > Maren > > -- Cornelius Werner cornelius.werner<at>gmail.com