there are a few things you could try:
- the command line tool 'avwroi' lets you extract an ROI, just try to
cut off the brain stem by removing a few slices.
- if removing a few slices does not do the trick, you could use flirt
to register the brain to standard space, apply the inverse to e.g.
avg152T1_brain_mask (in $FSLDIR/etc/standard ), dilate the new mask
(avwmaths in -dils out)
and use this to mask the original data (avwmaths in -mas mask out)
- optimising the BET parameters: the -g option lets you specify a
vertical gradient for driving the stimation (use negative values gives
smaller estimates at the brainstem. Also, if the brain is not well
located at the centre of the FOV, you can use the -c option to specify
the initial sphere.
hope this helps
On 19 May 2004, at 00:26, Russ Poldrack wrote:
> hi - we're having some trouble getting BET to work properly with
> MP-RAGE data from a Siemens Allegra. The problem is that the carotids
> are very bright, and BET consistently includes a chunk of neck tissue
> within the segmented brain, leading to bad registration errors with the
> standard for some subjects. We have tried a wide range of BET
> parameters but have not been able to prevent this. We've solved the
> problem so far by trying to optimize BET for the rest of the brain and
> then removing the offending sections by hand, but clearly it would be
> nice if BET could deal with these images more adequately. Do you have
> any suggestions about tricks that we might try (other than getting a
> new anatomical sequence)?
> Russell A. Poldrack, Ph.D.
> Assistant Professor of Psychology, UCLA
> Franz Hall, Box 951563
> Los Angeles, CA 90095-1563
> email: [log in to unmask]
> phone: 310.794.1224
> fax: 310.206.5895
> web: http://www.poldracklab.org/
Christian F. Beckmann
Oxford University Centre for Functional
Magnetic Resonance Imaging of the Brain,
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
Email: [log in to unmask] - http://www.fmrib.ox.ac.uk/~beckmann/
Phone: +44(0)1865 222782 Fax: +44(0)1865 222717