Hello Afni Experts,
I was interested in warping an MNI ROI to subject (EPI) space. I used 3dQwarp to create a deformation field and concatenated all of the rigid/affine/non-linear files into one warp field using 3dNwarpCat. This warp file is EPI2MNI+tlrc and takes my warp into EPI resolution. This part worked out perfectly. I wanted to then pass this warp along to 3dfractionize to specify the warp from MNI to EPI subject space, this way I can also use -clip etc as some of my ROI’s otherwise leak into adjacent CSF regions when directly using 3dNwarpApply.
I tried the following:
a) 3dfractionize -template epiref.align+orig -input ROI1+tlrc -warp EPI2MNI+tlrc -vote -clip 0.2 -prefix ROI1
I keep getting an error : Warp dataset doesn’t contain a warp transformation!
b) I also tried used 3dNwarpApply to apply the EPI2MNI+tlrc warp to epiref.align+orig to create a warped version which could potentially supply the warp field.
3dfractionize -template epiref.align+orig -input ROI1+tlrc -warp epiref.align+tlrc -vote -clip 0.2 -prefix ROI1
I keep getting an error : Warp dataset doesn’t contain a warp transformation!
When using the warp field directly or a warped input it does not seem to work. Previous posts suggest using @auto_tlrc to linearly warp the ROI, however this would not really help in my case as I want to take advantage of the nonlinear warp field for more accurate ROI placement. Is there a way to circumvent this error or am I missing another step in order to take advantage of 3dfractionize’s -warp option with a nonlinear field? Or is this function limited to just affine transformation matrices?
Thanks,
Ajay