Hi,
I think your approach looks reasonable - but the next version of
randomise deals with confounds in a more correct way - so I suggest
re-running randomise once the new release is out - very soon!
Cheers.
On 7 Aug 2007, at 22:01, Brian Schweinsburg wrote:
> Dear TBSS/randomise Users,
>
> We have a 1 group design (n=76) and would like to use a cognitive
> score to
> predict FA in the TBSS maps. We also have two confound EVs (age &
> biomarker
> data). The EVs have been demeaned. Our goal is to evaluate the
> relationship
> between cognition and FA while regressing out the age and biomarker
> data. I
> would greatly appreciate if someone experienced with creating
> design and
> contrast matrices could examine our approach and provide some
> feedback. In
> other words, do our goals and matrices match up appropriately. I am
> suspicious that I did something wrong or am not understanding the
> approach
> correctly. We have used the following command (machine = intel mac
> 10.4.10,
> 8 core; most recent version of FSL):
>
> randomise -i all_FA_skeletonised -o pat-globalNP-2 -m
> mean_FA_skeleton_mask
> -D -d design.mat -x confounds.mat -t design.con -n 5000 -c 2 -V
>
> **********
> design.mat
> **********
> /NumWaves 1
> /NumPoints 76
> /PPheights 1 1
>
> /Matrix
> 48.4666667
> 49.1333333
> 47.1333333
> 43.4
> .
> .
> .
> ***********
>
> *************
> confounds.mat
> *************
> /NumWaves 2
> /NumPoints 76
> /PPheights 1 1
>
> /Matrix
> 42 12.2474487
> 41 4.24264069
> 43 13.4164079
> 40 8.60232527
> 43 13.0384048
> 50 14.1421356
> .
> .
> .
> **************
>
> **********
> design.con
> **********
> /NumWaves 1
> /NumContrasts 2
> /PPheights 1 1
>
> /Matrix
> 1
> -1
> ***********
>
> Thank you for your valuable time.
> Brian
------------------------------------------------------------------------
---
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
------------------------------------------------------------------------
---
|