open-discussion > brain masking, skull striping
Showing 1-3 of 3 posts
Display:
Results per page:
Jan 4, 2018  03:01 PM | Daniel Berge - Hospital del Mar Medical Research Institute (IMIM)
brain masking, skull striping
Dear DTIPrep experts,
  I am trying to run brainmasking from the DTIPrep Pipeline modifying the defaults. It turns that braimaksing does not seem to strip skull. Using the "Slicer" flag it seems to run ok according to QCReport, but image QCed still has skull. Using any of the FSL options ends up with image math failure in the QCReport.
  Any solution? I've checked the protocol paths and everything seems to be where it should be . Usually I leave the "Brainmask_MaskedImage" field empty but I also have added a subject-specific mask in this field and it does the same.

Thanks in advance.
Jan 5, 2018  08:01 AM | Martin Styner
RE: brain masking, skull striping
The Slicer brain masking is not the best (which is likely why there's still skull in the image,I assume that something was removed?). For the FSL option, our tool ImageMath is used. Unfortunately there are other tools that have also an ImageMath binary (e.g. ANTs or Rview), so you need to make sure that the path to DTIPrep's ImageMath folder has to be before the paths of the other tools.

I usually use bet with IDWI

Martin
Jan 18, 2018  02:01 PM | Daniel Berge - Hospital del Mar Medical Research Institute (IMIM)
RE: brain masking, skull striping
Hi,
  After troubleshooting some issues with the paths of the tools, it turns that brainmask does not work because he cannot find subj_IDWI.nrrd. I noticed that when I check "compute the DTI" (usually I uncheck this last step), a subj_IDWI.nrrd file is generated. Does it make sense?

 Also, when trying to do visual inspection, it only saves the changes to a subj_QCed_VC.nrrd file if the DTI tensor had previously been computed. Does that make sense also or it is just something related to my computer?

Thanks