Hi Clare,
If you are unfamiliar with the threshold option, why do you want to
use it?
We no longer recommend this method. The "auto" option for boundary
correction via the run_first_all script is what we recommend as this
is then
tuned for each structure and generally uses the "fast" option. This
is also
the default behaviour if you do not specify the -m flag in
run_first_all.
If you really want to use the threshold method for some reason, then the
threshold values are z-thresholds applied to a normalisation of the
intensities
to form standardized z-score (accounting for mean and standard deviation
of the intensities within the structure).
All the best,
Mark
On 30 Jul 2010, at 17:06, Clare Gibbard wrote:
> Hi FSL experts,
>
> I would like to use the thresh option for FIRST boundary correction,
> but I wanted to confirm what threshold value I need to be specifying
> with the -t option. Is there a scale of thresholds which would
> usually be applied or should I be applying a threshold which is
> related to the intensity of my scans ie. a percentage of the mean
> intensity or something?
>
> I really appreciate your advice.
>
> Thanks very much,
> Clare
>
|