Hi,

I would start by reading the randomise manual carefully to make sure you understand the different files output by randomise, then have a careful look at these in FSLView to make sure that you are actually getting significant results (>0.95 in the _corrp file). You can extract location information about significant clusters using the "cluster" program.

Cheers.



On 27 Mar 2009, at 18:05, Yapeng WANG wrote:

Hi TBSS experts,

I used TBSS to analysize my DTI data and got TBSS stat. results. Right now I'd like to

present or extract FA values in a standard space (say MNI 152) with a .txt or excel file

and correlate FA values with behavioral results. But I do not how to do this.

I tried to do that based on Gwenaelle's suggestion. S/he told me I can run the following

scripts:

1) fslmaths my_tbss_results_corrp.nii.gz -thr 0.95 -bin my_significant_results_mask

then:

2) fslmeants -i all_FA_skeletonised -m my_significant_results_mask -o

my_significant_results.txt

After I run these two scripts, I got 2 files: my_significant_results_mask. nii.gz and my_significant_results. txt.

In the .txt file, when I run my tbss__results_corrp.nii.gz, the .txt file shows six 0s in a

column; and when I run my tbss_results.nii.gz, the .txt file shows six numerical values

ranged from .4000 to .7000. So I guess they are  FA values that I expect, but there are

no standard coordinates.

Can anyone tell me how to know the standard coordinate corresponding a specific FA

value? Ideally, I'd like to know the FA values and corresponding standard coordinates

from two groups of our subjects. Thank you!

yp

---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------