Hi,
This certainly is odd and looks wrong.
It might be that there was a problem in creating the image for the webpage rather than with the actual data, given that the final registration seems OK, although it is possible that it is doing it without the right unwarping but most places look relatively OK in this static view.
To check this it is best to load the images directly from the FEAT directory into a viewer (FSLeyes or FSLView). The images to look at are:
reg/unwarp/FM_UD_fmap_mag_brain2str
reg/highres
These should be well aligned. If they also look like what you have in the report page then there is a problem that needs fixing. It might be due to brain extracted versions of some images not being found - so check that you've entered the right images and always have a naming convention of somename.nii.gz and somename_brain.nii.gz (where obviously "somename" is whatever you want). If you have both of these (the non-brain-extracted and brain-extracted versions) in the same directory then things should work OK.
All the best,
Mark
> On 3 Sep 2017, at 17:02, Ricky Lin - McGill <[log in to unmask]> wrote:
>
> Hello FSL experts, I have a small question about BBR registration using fieldmap correction. I have checked that the T1-image is properly stripped, fieldmap mag and phase are processed properly. My question is on the registration report page. As shown in "Registration of fieldmap to highres" there seems to be a shift/translation on the alignment.
>
> https://photos.app.goo.gl/neYpj7JSn4DP5QUw1
>
> Is this normal/acceptable? It seems that the example_func2highres registration with fieldmap correction is doing reasonably well though.
>
> Thank you!
> Ricky~
|