Hi Liang,
you need to use the monkey analysis stream as many of the steps simply
don't apply to nonhuman data (such as the automated Talairach alignment!)
cheers
Bruce
On
Wed, 17 Nov 2010, liang wang wrote:
Hi Bruce,
I had tried to run recon-all on the dicoms. But the same errors show up.
Hi FSusers,
The terminal displays errors when I run recon-all -autorecon1 -subjid ...
to get a skull-stripped brain image. The processing exits with the following
errors regarding talairach transformation. Here is the talairach.xfm:
MNI Transform File
% tkregister2
Transform_Type = Linear;
wow, that's a completely wacky talairach xform. Did you run this straight
on the dicom images? The talairach clearly failed for some reason. Try
running mri_info on the dicoms to see if there is something strange in
them
On Wed, 17 Nov 2010, liang
wang wrote:
Hi FSusers,
The terminal
Hi Bruce,
I run the command on a NIFTI image converted by dicom files. The image can
be skull stripped using bet in FSL, but some skulls were not removed
clearly. Here is the information by mri_info. It seems to me normal, but
several scan parameters loss. For the nifti image, I use to3d (AFNI
can you try running recon-all directly on the dicoms? Also, it's probably
pretty
noisy at .5mm in-plane - how does it look? Is it whole brain?
On Wed, 17 Nov
2010, liang wang wrote:
Hi Bruce,
I run the command on a NIFTI image converted by dicom files. The image can
be skull stripped
Hi Bruce,
I had tried to run recon-all on the dicoms. But the same errors show up.
This image covering whole brain is obtained (Siemens 3T) actually from a
monkey and looks good. Here is the header information from the dicom image.
According to the suggestion in Freesurfer FAQ, I run tkregister2
