External Email - Use Caution
Hi!
Thanks, Doug and Bruce, for your input.
It seems to be a problem with the number of iterations (-n). It is
currently hard coded to run 4 times. If I run it only once, grey and white
are in a normal range (but there are still inhomogeneities). If I r
It's been a while since 5.3, but I vaguely remember something like this
happening because there were one or two super bright voxels in the
image. If this is the case, then you can find them and clip them.
On 9/18/20 10:32 AM, Fischl, Bruce wrote:
I don’t think adding 100 will work as once it
I don’t think adding 100 will work as once it scales everything down that far
you only have a couple of bits of representation in the conformed ( 1
byte/voxel) images. Maybe someone else can comment on fixing nu_correct?
From: freesurfer-boun...@nmr.mgh.harvard.edu
On Behalf Of Caspar M. Schwi
Hello Teodora,
In version 5.3 of freesurfer, The Talairach transform step will fail at
the mri_nu_correct.mni command on newer Linux distributions that come
with perl version 5.22 (and higher). This is due to an incompatibility
with the mni tools shipped with FreeSurfer v5.3 and newer versions
Hi Falk,
glad to hear that I am not the only one, and thanks!
Actually I cropped the image with mri_convert --cropsize to a size wich has one
voxel less in every dimension, and it worked.
Not sure, but maybe cropping is better if one wants to avoid interpolation
(which, in turn, can introduce
Hi Natalia,
I had the same issue in some cases. This is happens if you native data is not
perfectly isotropic and seems to be related to a rounding error. The only the
solution I found is to fix the resolution of your data prior to mri_nu_correct
manually.
Check your voxel size with mri_info.
Hi Cristian and Anna
just do it outside of FS. Convert the volume to minc with mri_convert and
run the MNI nu_correct on it directly
cheers
Bruce
On Thu, 11 Jul 2013, Cristian Paz wrote:
Hi Anna and all Freesurfers,
I have sort of the same problem.
I know there is a flag called --noconf
Hi Anna and all Freesurfers,
I have sort of the same problem.
I know there is a flag called --noconform - but it doesnt seem to work on
my data with mri_nu_correct.mni . (does not recognize the flag )
I also tried the flag --cm which is supposed to only minimally conform -
but here I also got a
Hi Cristian
you can use mri_normalize directly on the T1, but it won't work for the
T2 unless you specify control points and tell it to only use them. There
are other (more esoteric) options in mri_normalize that you could use.
Look at the recon-all code for the T2-based dura correction
chee
Hi Nikolas,
did you see the "no space left on device" error? Is your partition full?
I can convert the orig.mgz with no problem. Also, why is the image 512^3?
That's a giant image, and at 1mm way more than you need to fit a human
head. Did you zeropad it for some reason?
cheers,
Bruce
On Wed,
Hi Jeff,
sorry, I don't know what causes the nu_correct to do this, but you'll
probably have to skip that step as otherwise you've lost too much dynamic
range. I would try copying orig.mgz to nu.mgz and running all the steps
afterwards.
cheers
Bruce
On Wed, 7 Oct 2009, Jeff Sadino wrote:
> He
11 matches
Mail list logo
