Hi Gaby,

It sounds like you have so many subjects that at the 1x1x1mm  
resolution (that the later TBSS stages are run at), you are generally  
going to have RAM problems unless you can get access to a 64-bit  
machine.

I'm assuming that the mean_FA_skeleton was created ok by the call to  
tbss_skeleton in tbss_3_postreg. The second call, the one you are  
referring to in tbss_4_prestats, takes in the 4D data all_FA and  
creates the skeleton-projected 4D output file all_FA_skeletonised.  
You are correct that it would be fine to replace this call with calls  
to subsets of the 4D all_FA data, and then remerge the output. You  
can split files with avwroi (maybe try avwroi++) and remerge with  
avwmerge++

Thus you should be able to end up with a recombined  
all_FA_skeletonised. However, you do need this complete file in order  
to feed it into randomise - you can't process subsets (in the 4th =  
subject ID dimension) of that separately for obvious reasons. You  
_could_ take subsets spatially - e.g. use avwroi to split the top and  
bottom half (in z) - but then randomise can't do multiple comparisons  
across space properly. It might be more realistic to decide on a  
central cuboid of the data that contains all areas of interest and  
just run randomise on that.

Hope this helps - cheers, Steve.





On 18 Apr 2006, at 04:06, Gaby Pell wrote:

> Hi Steve,
>
> Thanks for the advice. avwmerge++ worked a treat on successfully  
> merging
> the files - but tbss_4_prestatsis is now complaining at the  
> tbss_skeleton
> stage. For smaller collections of images, it works OK - so it must  
> be the
> same issue. Any suggestions?
>
> If I ran tbss_skeleton on multiple sub-groups of the 4D data set to
> project the data set on the skeleton in smaller chunks, is there  
> any way
> to run the randomise script in a way that will group these blocks
> together.
>
> Thanks,
>
> Gaby


------------------------------------------------------------------------ 
---
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
------------------------------------------------------------------------ 
---