Hi Gaby,
On 17 Apr 2006, at 07:39, Gaby Pell wrote:
> Thanks for making the tbss package available.
>
> Unfortunately, I am getting core dump error messages during the
> avwmerge
> in tbss_3_postreg. This has always been on an issue on our
> computers when
> merging more than 50 images or so (memory?). I am working currently
> with
> around 100 FA images. Is there any way around this?
You could check that this doesn't get solved by increasing the amount
of swap, but I suspect that you've already tried this. Presumably the
problem would also go away if you used a 64-bit computer.
Finally (and probably most interestingly to you), you might try
editing the tbss script to call avwmerge++ instead of avwmerge - this
is a beta-test c++ recoding of avwmerge which I think uses less
memory than the original - would be interesting to know if this
solves the problem.
> I forgot to add that I could bypass the all_FA merge and create the
> mean_FA image by simply adding and dividing. However, I can see that
> tbss_4_prestats uses the all_FA image. Could that final process
> simply be
> split into a number of sub-blocks of images or does it need to be
> in one
> go?
You do need the merge - as well as producing a mean FA image and then
the skeleton of this, TBSS then goes on to project the 4D data onto
this skeleton.
Cheers, Steve.
------------------------------------------------------------------------
---
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
------------------------------------------------------------------------
---
|