Re: [HCP-Users] most typical subject for FA registration?

2017-04-01 Thread Glasser, Matthew
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,

Re: [HCP-Users] most typical subject for FA registration?

2017-04-01 Thread Glasser, Matthew
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:

Re: [HCP-Users] most typical subject for FA registration?

2017-04-01 Thread John Griffiths
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

[HCP-Users] [fMRIVolumeProcessingPipeline]cannot find eye.dat

2017-04-01 Thread --
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

Re: [HCP-Users] [fMRIVolumeProcessingPipeline]cannot find eye.dat

2017-04-01 Thread Glasser, Matthew
The order of running the pipelines is: PreFreeSurfer —> FreeSurfer —> PostFreeSurfer —> fMRIVolume —> fMRISurface —> ICA+FIX —> MSMAll PostFreeSurfer —> DiffusionPreprocessing —> BedpostX Peace, Matt. From: >

[HCP-Users] most typical subject for FA registration?

2017-04-01 Thread John Griffiths
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)