hi steve, thanks for your answer ! in the meanwhile, I actually did find a
reasonably easy hack for feat.tcl, which will work if you specify a
brain/background threshold of 0%...
at line 5549 on the file feat.tcl, use the following:
if { $fmri(brain_thresh) == 0 } {
# the following line is the original version
# set absbrainthresh [ fsl:exec "$FSLDIR/bin/avwstats ${funcdata} -R |
awk '{ print \$1 }' -" ]
# instead, do the following:
# step1: compute the threshold as being the min value of the
filtered_func_data, instead of using the original data (which I think might
be a bug anyway; not sure, I didn't look too much into that)
set absbrainthresh [ fsl:exec "$FSLDIR/bin/avwstats filtered_func_data
-R | awk '{ print \$1 }' -" ]
# step 2: substract 1 to that value (in case this threshold is used in a
strict inequality by the following programs, which I didn't check):
set absbrainthresh [ expr $absbrainthresh - 1 ]
# step 3: add 1 to the mask (previously created by ip), to make sure
that values in the mask are non-zero everywhere
fsl:exec "${FSLDIR}/bin/avwmaths mask -add 1 mask"
}
it's not very clean, but it does the job; that can be useful to work on data
with an non brain-like histogram: simulated data for instance, or spatially
high-pass filtered versions of normal brain images...
cheers,
sylvain
On Tue, 4 Nov 2003 05:59:29 +0000, Stephen Smith <[log in to unmask]> wrote:
>Hi Sylvain,
>
>Yep, you're correct, FEAT assumes you're putting a brain image into it ;-)
>
>Probably the simplest workaround is just to pad with an extra slice, with
>values set to lower than those you want to keep, using avwmaths and
>avwmerge etc....i've looked in the FEAT tcl and it's not trivial to change
>there - the other workaround would be a small edit in ip.c and to
>recompile that.
>
>Cheers, Steve.
>
>
>On Sun, 2 Nov 2003, Sylvain Takerkart wrote:
>
>> hi all !
>>
>> I'm struggling to force feat to be applied on ALL voxels in my data
>> volume; the call to 'ip' defines a threshold, and a mask; even if you
>> select 0% threshold, there's an adjustment up to 2% (according to what
>> I've seen in the source code for the threshold selection) which removes
>> some voxels...
>>
>> is there any simple way to do that ?
>>
>> thanks,
>>
>> sylvain
>>
>
> Stephen M. Smith DPhil
> Associate Director, FMRIB and Analysis Research Coordinator
>
> Oxford University Centre for Functional MRI of the Brain
> John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
> +44 (0) 1865 222726 (fax 222717)
>
> [log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|