No, concatenate the inverse of it on to the end of each subject’s registration
to remove the registration bias without having to do an enormous amount of
pairwise computations.
Peace,
Matt.
From: John Griffiths
>
Date: Saturday,
That does seem like a lot of computations. Why not pick a random individual,
register to that and then remove the group average effect of registration,
retaining the improved alignment across subjects while eliminating the bias
caused by the initial subject?
Peace,
Matt.
From:
Not sure exactly what you mean there by removing the group average effect.
In the design matrix?
I am now doing effectively that with a random subject, just wanted to draw
/ leech on the community to see if anyone was able to make an informed
recommendation about which subject to choose for the
Dear professors,
I tried to run the
Pipelines-3.4.0/Examples/Scripts/GenericfMRIVolumeProcessingPipelineBatch.mine.sh
(I modified based on HCP course practicals), and I got an error when
Pipelines-3.4.0/fMRIVolume/scripts/DistortionCorrectionAndEPIToT1wReg_FLIRTBBRAndFreeSurferBBRbased.sh
The order of running the pipelines is: PreFreeSurfer —> FreeSurfer —>
PostFreeSurfer —> fMRIVolume —> fMRISurface —> ICA+FIX —> MSMAll
PostFreeSurfer —> DiffusionPreprocessing —> BedpostX
Peace,
Matt.
From:
>
Hello.
I was wondering if anyone has done the all-to-all TBSS registration step on
FA images from the WU-Minn data to find the most typical subject for
two-step FNIRT FA registration?
If so would it be possible to get the most typical subjects' IDs?
This would save me (and I'm sure many others)