Dear Anderson,
Thank you for your reply!
The test that I will be performing is a permutation test, this is how
the original p-value image is obtained. Do you mean I should run TFCE
using fslmaths at each permutation? Can I then use fslmaths to combine
the TFCE outputs produced at each iteration?
Can you please give me a bit more detail or send me a link to an
appropriate part of the fslmaths user manual? The only (vaguely)
related guidance that I found is this:
http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Cluster
Thank you very much in advance!
Best wishes,
Kate.
Quoting "Anderson M. Winkler" <[log in to unmask]> on Tue, 6 May
2014 23:21:30 +0100:
> Hi Kate,
>
> So you did a test, produced a p-value, and would like now a TFCE image,
> which of course can only be interpreted together with a TFCE p-value.
> However, a TFCE p-value can currently only be produced via permutations. If
> you can do some kind of (valid) permutation in your test, a TFCE for each
> permutation can be produced with fslmaths. A little bit of code lets you
> calculate the p-value then.
>
> All the best,
>
> Anderson
>
>
>
> On 6 May 2014 13:52, Ksenia Andreyeva <[log in to unmask]> wrote:
>
>> Dear FSL,
>>
>> I am writing code for a different type of analysis than the one available
>> using fsl randomise, it is also performed on voxel-wise basis using TBSS
>> skeletons, producing as a result a p-value image. I would like to use the
>> TFCE tool from randomise to correct for multiple comparison.
>>
>> Could you please tell me how I can do that separately from the full TBSS
>> randomise analysis just on a p-value image?
>>
>> Thank you very much in advance!
>>
>> Best wishes,
>> Kate Andreyeva.
>>
>>
>> --
>> The University of Edinburgh is a charitable body, registered in
>> Scotland, with registration number SC005336.
>>
>
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
|