Dear experts,
I have run featquery based on masks created after group analyses. Two of the clusters are fairly small (randomize with TFCE on functional data). These clusters end up being fewer than 10 voxels in native space (~2x2x4 resolution). On several occasions, the transformation into native space failed completely. Interestingly, it only fails for one of the 2 functional runs (2 different conditions during the same imaging session). In my pipeline, I register the anatomical to the template once and register each functional to the same anatomical. All the registrations look fine. What are the possible reasons for this behavior?
On a related note, I'm curious why the inv_warp.nii and the warp.nii files look so different. The warp file is pixelated while the inv_warp file is smooth. Is that normal?
Thank you in advance for your help.
-Marina
|