>From previous post:
There are four rfMRI runs per subject. LR vs RL refers to the phase encoding
direction and 1 and 2 refers to the session each run was acquired in. You
should use balanced amounts of RL and LR data ideally, as the phase encoding
direction does create some asymmetry in the data. You want to use the
MSMAll_hp2000_clean dtseries file, as this file has its cortical areas aligned
across subjects. If you wish to concatenate across runs, you should remove the
mean image. You could also variance normalize the data with by multiplying by
the bias field and dividing by the variance normalization map though these maps
would need to be resampled into MSMAll alignment as the currently distributed
files are in MSMSulc alignment. This can be done using:
#Create spheres for resampling from MSMSulc to MSMAll on the 32k mesh
wb_command -surface-project-unproject
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/Native/${Subject}.L.sphere.MSMSulc.native.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/Native/${Subject}.L.sphere.MSMAll.native.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
wb_command -surface-project-unproject
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/Native/${Subject}.R.sphere.MSMSulc.native.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/Native/${Subject}.R.sphere.MSMAll.native.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
#Resample Variance Normalization Map
wb_command -cifti-resample
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_hp2000_clean_vn.dscalar.nii
COLUMN
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_hp2000_clean_vn.dscalar.nii
COLUMN ADAP_BARY_AREA ENCLOSING_VOXEL
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_vn.dscalar.nii
-surface-postdilate 30 -nearest -left-spheres
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.32k_fs_LR.surf.gii
-left-area-surfs
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.L.midthickness.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.L.midthickness_MSMAll.32k_fs_LR.surf.gii
-right-spheres
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.32k_fs_LR.surf.gii
-right-area-surfs
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.R.midthickness.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.R.midthickness_MSMAll.32k_fs_LR.surf.gii
#Resample Bias Field Map
wb_command -cifti-resample
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_hp2000_clean_bias.dscalar.nii
COLUMN
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_hp2000_clean_bias.dscalar.nii
COLUMN ADAP_BARY_AREA ENCLOSING_VOXEL
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_bias.dscalar.nii
-surface-postdilate 30 -nearest -left-spheres
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.32k_fs_LR.surf.gii
-left-area-surfs
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.L.midthickness.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.L.midthickness_MSMAll.32k_fs_LR.surf.gii
-right-spheres
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.32k_fs_LR.surf.gii
-right-area-surfs
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.R.midthickness.32k_fs_LR.surf.gii
${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.R.midthickness_MSMAll.32k_fs_LR.surf.gii
#Remove spheres from above
rm
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.L.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
rm
${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.R.sphere.MSMSulc_MSMAll.32k_fs_LR.surf.gii
#Create mean image for demeaning
wb_command -cifti-reduce
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii
MEAN
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_mean.dscalar.nii
#Demean and variance normalize timeseries (and revert bias field correction)
wb_command -cifti-math “((TCS - Mean) * Bias) / VN”
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_nobias_vn.dtseries.nii
-var TCS
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean.dtseries.nii
-var Mean
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_mean.dscalar.nii
-var Bias
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_bias.dscalar.nii
-var VN
${StudyFolder}/${Subject}/MNINonLinear/Results/${fMRIName}/${fMRIName}_Atlas_MSMAll_hp2000_clean_vn.dscalar.nii
Peace,
Matt.
From: HERACLES PANAGIOTIDES <he...@uw.edu<mailto:he...@uw.edu>>
Date: Monday, January 1, 2018 at 6:26 PM
To: Matt Glasser <glass...@wustl.edu<mailto:glass...@wustl.edu>>
Subject: Re: [HCP-Users] (no subject)
Hi Matt and Happy New Year.
Is the processing already done and available on the site? If o, what are the
processed files called?
Thanks,
Heracles
From: Glasser, Matthew
Sent: Sunday, December 31, 2017 9:35 AM
To: hcp-users@humanconnectome.org
Cc: HERACLES PANAGIOTIDES
Subject: Re: [HCP-Users] (no subject)
The variance normalization is per timeseries. We prefer to use the _vn files
provided by the RestingStateStats pipeline so that this is equalizing the
unstructured noise.
Peace,
Matt.
From: HERACLES PANAGIOTIDES <he...@uw.edu>
Date: Saturday, December 30, 2017 at 2:05 PM
To: Matt Glasser <glass...@wustl.edu>
Subject: Re: [HCP-Users] (no subject)
Hello Matt,
I am getting ready to launch a group analysis of rfMRI files using Matlab. I
want to make sure that I am not making any fundamental errors in preprocessing.
I would be deeply appreciative, if I could (phone?) consult with you or anyone
else, to make sure that I am not making any gross mistakes. At this point, I
am following your instructions about mean and and variance normalization before
ROI extraction. For example, is this equivalent to a z-score transformation?
If so, is this done to the entire 4D volume or to the individual voxel time
series?
At any rate, these are some of my questions.
Thank you for being so kind.
Heracles
_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
