Hi,
On 12 Jul 2006, at 22:44, iris bautista wrote:
> Hello FSL Users,
>
> I have been using BET v1.2 in FSL 3.3 on Linux Redhat9 to strip my
> structural images that will be used for VBM analysis but several
> problems occured.
>
> 1) First I kept the fractional threshold at 0.5, but some non-brain
> tissue remained (as seen on
> https://webspace.utexas.edu/ieb58/MPRAGE1-CG1013_brain.jpg). It was
> eliminated after changing the threshold to 0.6 but I lost too much
> brain volume. Is it better to leave it at 0.55 even with the extra
> tissue to preserve brain volume? I'm not sure if the extra tissue
> would result in innacurate data in my analysis.
>
> 2) Since changing the threshold did not work, I used avwroi to crop
> the original image which resulted in a decent extraction. However,
> after overlaying the extracted image on the original, part of the
> cerebellum was lost (as seen on
> https://webspace.utexas.edu/ieb58/avMPRAGE1-CG1013_overlay2.jpg). How
> much brain volume loss is acceptable when studying structural images?
Instead of using avwroi you could also use the -c option to tell BET
where to center the starting estimate - this has the advantage of
still being in the original space then. On the other hand, when you
crop an image to remove unnecessary slices at the bottom, that's
generally a good thing to do as you rarely want to keep them anyway.
On your cropped image have you tried using the -f and the -g options?
Looks like you should be able to get a bit of an improvement, though
it does look pretty good already.
> 3) I still used the image from step 2 on SPM5 in MATLAB 7.0 to see if
> it would cause any problems. I first used the "check register" button
> to check the alignment of the stripped template (using BET with 0.5
> threshold) with the image from step 2. The images, however, do not
> align even after I reoriented the extracted brain. I'm not sure what
> else to do at this point.
You can try the above and see if that fixes this too - it may be that
this problem relates to the setting of a coordinate origin.
Cheers, Steve.
>
> Thank you for your help!
------------------------------------------------------------------------
---
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
------------------------------------------------------------------------
---
|