open-discussion > DTIPrep movement estimation output?
Showing 1-3 of 3 posts
Display:
Results per page:
Jan 25, 2018  05:01 PM | Ryan H - UCLA
DTIPrep movement estimation output?
Hi all,

I'm wondering if I can retrieve estimated movement values between volumes from DTIPrep. FSL eddy automatically outputs a text file indicating between-volume movement for each subject along with cumulative movement estimates relative to the first volume. I use these estimated movement values to exclude certain subjects from analysis.


Thanks!
Jan 25, 2018  05:01 PM | Ryan H - UCLA
RE: DTIPrep movement estimation output?
I've found that this movement output is labeled "TranslationNorm" in the *XMLQCREsult.xml output file. However, I see a value of 0 for each volume, implying there was absolutely zero movement between each volume, which seems implausible. It may help to know that I am using DTIPrep v1.2.4 and Slicer 4.6.2 (for when I use DWIConvert to go from NIFTI to NRRD).

Any help would be appreciated! Thanks again!
Jan 25, 2018  05:01 PM | Martin Styner
RE: DTIPrep movement estimation output?
Hi Ryan

"TranslationNorm" is indeed the total amount of translative motion found by the motion correction (each DWI has such an entry). The amount of rotation motion is found in the output "Angle".

Keep in mind that this is only the translation and rotation recovered by the motion & eddy current correction step in DTIPrep. If you ran eddy with FSL first, that information is not used/accessible by DTIPrep.

If you ran the motion & eddy current correction and get a TranslationNorm of exactly 0, then something is likely off. There should be some minor translation even just due to noise. Did you run the motion & eddy current step?

Best regards
Martin