Hi Ed,
This is a bug in randomise that results in output files containing "default" orientation values when no mask is supplied to randomise. A potential workaround here is to create an "all-ones" mask from a single volume of your input data.
Kind Regards
Matthew
> Hi,
>
> I'm reposting my message below:
>
>
> I have applied randomise to a 4D volume in Talairach space with an
> isotropic voxel size of 3 mm:
>
> randomise -i Volume4D -o ttest -T -d ../design.mat -t ../design.con -
> n 5000
>
> The output is a number of tstat files all having an isotropic voxel
> size of 1 mm instead of 3 mm and
> with no sform and qform information in the header. Is perhaps the
> fact that the 4D volume is in
> Talairach space causing these erroneous results?
>
> fsl version 5.0.4
> Mac OSX 10.6
>
> Ed
|