Hi,
I'm using a combination of FLIRT and FNIRT to register one brain to
another,
both T1 weighted, but the result is not good. I tried a number of
different
parameter settings without succes.
My commands are:
flirt -cost corratio -dof 12 -in myT1 -ref refT1 -omat refT1_to_myT1.mat
fnirt --in=myT1 --ref=refT1 --aff=refT1_to_myT1.mat --
cout=refT1_to_myT1_warp
applywarp -i myT1 -r refT1 -w refT1_to_myT1_warp -o refT1_to_myT1
with
myT1 is a skullstripped brain, voxel size 1 x 1 x 1 mm^3, matrix 256
cubic and
refT1 is a brain (not skullstripped) with the same voxel size, matrix
181 x 217 x 181.
Both are in standard (=axial) orientation.
The result is very weird, the orientation of the registered brain has
become
sagittal. My guess was that this was due to the fact that myT1 is
skullstripped
and refT1 not.
So I tried the same command but now with a skullstripped version of
refT1.
Although the orientation of the result matched that of the target
brain (myT1),
the result was not good: the registered brain looked like a blown-up
version
of the target brain. The intermediate result, i.e., the output of
FLIRT, fairly
matched the target brain, though.
Can you give me some advice on how to get a better result.
Cheers,
Ed
|