Aha. OK, omitting -no-resample, as below, outputs a well registered, but
downsampled, anatomical file. What should I change to output a well registered
anatomical file with the same dimension and voxel size as the original
anatomical? If this was in the mri_vol2vol help, it was not clear to
Thanks. The tkregister2 check gives some new information. The initial
registration is not very bad when checked in tkregister2. Also, the final
registration is perfectly acceptable when checked in tkregister2, so my
original question needs clarification.
What I want to do is to use the
Why are you using --no-resample? This will keep the anatomical in the
256^3 space (it just changes the header).
doug
Hansen, Kathleen (NIH/NIMH) [F] wrote:
Thanks. The tkregister2 check gives some new information. The initial
registration is not very bad when checked in tkregister2.
Thanks for the suggestions. It is whole-brain epi data. Unfortunately, it was
not acquired at the same time as the anatomical. Just for the heck of it I
tried substituting --init-header for --init-fsl anyway, and got the same
warning about negative G-W contrast.
-Kathleen
On 8/19/10 2:39
That warning in and of itself is not critical, but it does suggest that
the initial registration failed. I assume the registration still looks
bad? Does this happen on a bunch of data sets or just this one? There
are a couple of more steps. First, you can run with --init-fsl and check
the
Hello,
The following command has produced a very wrong registration and a warning:
bbregister --mov epi.template.strange_G-W_contrast.nii --bold --s D06
--init-fsl --reg register.strange_G-W_contrast.dat
WARNING: initial G-W contrast is negative, but expecting positive.
If the mov data has
It may have been the initialization that failed. Was the epi acquired
at the same time as the anatomical? If so, you might be able to run it
with --init-header. Also, is this whole brain or partial field of view?
doug
On 8/19/10 1:50 PM, Hansen, Kathleen (NIH/NIMH) [F] wrote:
Hello,
The
