Sure - if you don't intend to take these results up to higher-level
(across subjects) then you can make the masking more conservative.
Edit $FSLDIR/tcl/featlib.tcl and:
change
fsl:exec "${FSLDIR}/bin/bet2 mean_func mask -f 0.3 -n -m; ${FSLDIR}/
bin/immv mask_mask mask"
to have a higher -f option in BET
and maybe delete the line
fsl:exec "${FSLDIR}/bin/fslmaths mask -dilF mask"
which dilates the mask
Cheers.
On 5 Oct 2008, at 02:15, Gui Xue wrote:
> Thanks a lot, Joe and Steve.
> I wonder if there is a way to play with the parameters so that we
> can have a simple fix?
>
> cheers,
> Gui
>
>
> On Thu, Oct 2, 2008 at 1:50 PM, Joe Devlin <[log in to unmask]>
> wrote:
> Hi Gui,
>
> The motivation comes from a discussion we had on the list a few
> months (maybe more?) ago. FSL 3.1 used a more restrictive masking
> procedure that ended up losing voxels that were clearly inside the
> brain. A few of us complained that this was overly conservative and
> that the original masking approach was better precisely because we
> don't lose intra-brain voxels. The cost is the potential for false
> positives with activation outside the brain, although in my
> experience, this was not a serious problem, at least not at a group
> level. I take your point, though, that it is labor intensive to
> fix such voxels if they form significant parts of clusters, etc.
>
> I think Steve and the FSL team are getting conflicting feedback --
> too restrictive or too lenient? Steve if you guys could just get it
> perfect, the whole problem would just go away! :)
>
> Joe
>
---------------------------------------------------------------------------
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
---------------------------------------------------------------------------
|