HI - no, not quite. You should:
- subtract the average of EV2 from it (i.e. "demean" it).
- set the contrasts to [0 1] and [0 -1].
Steve.
On 22 Dec 2008, at 21:46, SUBSCRIBE FSL takeshiasami wrote:
> Dear FSL users,
>
> I'm a beginner of TBSS(v1.2) and now planning to do correlation
> analysis
> between FA values and clinical measurements (positive and negative
> correlation) in the patients group. I learned how to do it from this
> year's
> FSL mailing list, but I don't have enought confidence. Please check
> my method.
>
> 1. creating skeletons (mean_FA, all_FA, etc) with 5 patients.
> 2. creating desgin matrix and contrast using Glm window.
>
> [EVs]tag
>
> Number of main EVs: 2
> Number of additional, voxel-dependent EVs: 0
>
> Group EV1 EV2
> Input1 1 1 23
> Input2 1 1 14
> Input3 1 1 8
> Input4 1 1 36
> Input5 1 1 44
>
> *The numbers in EV2 column are clinical mesurement.
>
> [Contrasts & F-tests]tag
>
> Contrasts: 2, F-test: 0,
>
> Title EV1 EV2
> C1 positive correlation 1 1
> C2 negative correlation 1 -1
>
> 3. running randomise software
>
> randomise -i all_FA_skeletonised -o tbss -m mean_FA_skeleton_mask -d
> design.mat -t design.con -n 500 -D --T2 -V
>
>
> Is it OK?
>
> Sincerelly,
>
> Takeshi Asami
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|