Hi Jessica,
You are better off using native space because the boundary correction
works better on the original voxel intensities rather than ones that are
interpolated and resampled into standard space. This is why we do not
recommend the --useReconMNI option (in fact we will probably remove it
in the next version). There is an alternative for calculation that
does not
rely on boundary correction which Brian spoke about, but currently that
hasn't been implemented.
In the meantime you can transform the native volume values into standard
space values by multiplying them with the scaling factor that transforms
native to standard space. This can be found from the flirt
transformation
matrices with the avscale command as follows:
avscale vol2std.mat | grep Determinant | awk '{ print $3 }'
If you multiply the native space volumes (in mm^3) by this value then
you
will get the standard space volumes (in mm^3).
All the best,
Mark
On 22 Dec 2008, at 19:04, Jessica Kirkland wrote:
> Hi again,
>
> I apologize if I am asking a redundant question, but I am still in
> need of some clarification regarding FIRST segmentation. My goal is
> to compare FIRST volumes with hand-segmented volumes, which are in
> MNI space. After re-reading list emails, I am sure that my FIRST
> volume values are in native space, but I am still confused about my
> options for transformation. The last email from Brian said that the
> software for transforming volumes between spaces (the best
> transformation method) is not yet released. Is there another way, at
> another step, to get my volumes into MNI space? Another of the list
> emails indicated that for volume calculation with first_utils, I
> should be using --useReconNative (not --useReconMNI). Could someone
> explain why this is? Also, I tried the following script, which gave
> the following error. Can I get standard space structures in any way?
> Apologies if the answer is there and I am just missing it.
>
> Thank you for your time and patience!
>
> Jessica
>
> first_utils \
> --usebvars \
> -l 1 \
> -i 043_Lamyg_sup_con.bvars \
> -o Lamyg_supcon_vol \
> --useReconMNI \
>
> terminate called after throwing an instance of
> 'RBD_COMMON::BaseException'
>
> On Mon, Dec 15, 2008 at 1:42 PM, Jessica Kirkland <[log in to unmask]
> > wrote:
> Hi all,
>
> I am trying to get volume values for hippocampi and amygdalae
> segmented with FIRST, using fslstats -V. This option outputs 2
> values: voxels and volume. Are these values in native space?
>
> thank you,
>
> Jessica
>
