Try it in applywarp using --premat and -w (without the --rel flag). Does
that work better?
Peace,
Matt.
On 9/24/13 2:39 PM, "Courtney Haswell" <[log in to unmask]> wrote:
>Hi,
>
>I am trying to combine a FLIRT output and FNIRT output with convertwarp.
>Both of these work well on their own, but when combined with convertwarp,
>the output registration is very distorted and shifted in the space. Any
>idea what I am doing wrong? Thanks!
>
>Code:
>#Linearly register T1 to standard
>flirt -in T1_brain.nii.gz -ref MNI152_T1_2mm_brain -omat T1_FLT.mat -out
>T1_FLT -dof 12
>
>#Nonlinear registration of T1 to standard
>fnirt --in=T1.nii.gz --aff=T1_FLT.mat --cout=T1_FNT
>--config=T1_2_MNI152_2mm
>
>#Linearly register magnitude image
>flirt -in mag_brain -ref T1_FLT -out mag_FLT -omat mag_FLT.xfm -searchrx
>-60 60 -searchry -60 60 -searchrz -60 60 -dof 12
>
>#Combine magnitude linear matrix with T1 nonlinear matrix
>convertwarp --ref=MNI152_T1_2mm --warp1=T1_FNT.nii.gz
>--premat=mag_FLT.xfm --relout --out=Forward_warps
>
>#warp mag suscesceptibility to MNI
>applywarp --ref=MNI152_T1_1mm --in=qsm.nii --warp=Forward_warps --rel
>--out=qsm_FNT
|