I'm trying to create seed voxels (i.e., single-voxel masks) for featquery analysis by drawing them in FSLview. Featquery produces empty timecourses, and I believe this happens because my mask is in structural space (highres scan, not standard brain), while my data is in lower-res functional space. I tried to work around this by creating my masks from the functional images, but the low-res image makes it impossible to find the right anatomical landmarks. For science-logic reasons I do not want to create my mask in standard space.
As far as I can tell, the right solution would be to use applywarp to transform the masks from highres space into functional space. Here's what I tried, adapting the usually-useful instructions from the old "combining warps" FSL webpage:
invwarp --ref=example_func --warp=example_func2highres_warp --out=example_func2highres_warp_inv
applywarp --ref=example_func --in=MASK_HIGHRES --warp=example_func2highres_warp_inv --postmat=highres2example_func.mat --out=MASK_FUNC
Alas, this produces an empty output mask. Anyone have insights into what I might be doing wrong, and/or a better way to accomplish this?
Thanks!
-Benjamin Philip
University of Missouri
|