Dear Matthew,
I have updated for the new version 5.0.11.
Setup_masks creates me one 4D mask-file and design files with separate EVs for masks i.e. from 6....N
I run the suggested command:
randomise -i all_AD_skeletonised -o tbss_AD -m mean_FA_skeleton_mask -d masked.mat -t masked.con --T2 -V --vxl=6 --vxf=masked
after which I get the following warning:
"Warning: The unpermuted statistic image for the current image contains no positive values, and cannot be processed with TFCE. A blank output image will be created."
I get rid of the error if i instead run:
randomise -i all_AD_skeletonised -o tbss_AD -m mean_FA_skeleton_mask -d masked.mat -t masked.con --T2 -V --vxl=6,7,8...N --vxf=masked, masked,masked...masked
But I am not sure if the analysis is then run as I want. Probably the code is still expecting separate masks for each individual (as in the previous version=? But should this then be the 4D masked file or individual lesion mask.
Thank you in advance,
Noora
__________
Hello,
I would recommend using the latest version of FSL 5.0.11 for this - there’s been some minor tweaks in how setup_masks works with randomise.
Kind Regards
Matthew
--------------------------------
Dr Matthew Webster
FMRIB Centre
John Radcliffe Hospital
University of Oxford
> On 24 Apr 2018, at 08:13, Noora T <[log in to unmask]> wrote:
>
> Dear FSL experts.
>
> I updated my FSL version (5.09) after not being able to run randomise when using several masks as columns>rows.
>
> However, I am running into same problem as discussed previously:
> Problem: running randomise the unpermutated statistic is 0, generating a blank output file
>
> As the handeling of the masks in the new version was changed from several masks (mask1, mask2...etc) to just one 4D mask (masked), i wonder if the problem comes somehow now with design files. Similar design-files were still created (masked.con masked.mat).
>
> Has anyone so far figured out the problem as I could not find answer on the previous post?
>
> Thank you in advance,
> Noora
|