Functional data (general)
This covers general files that pertain to the preparation of the fMRI data.
nsddata/ppdata/subjAA/func1mm
This contains the high-resolution 1-mm preparation of the fMRI data.
nsddata/ppdata/subjAA/func1pt8mm
This contains the standard-resolution 1.8-mm preparation of the fMRI data.
nsddata/ppdata/subjAA/func*/brainmask.nii.gz
The binary brain mask that was used to mask the betas (in order to save disk space). Note that this brain mask is intentionally liberal so as to not lose brain voxels.
nsddata/ppdata/subjAA/func*/meanBBBB.nii.gz
nsddata/freesurfer/subjAA/label/[lh,rh].mean.mgz
This is the mean of all of the pre-processed volumes in BBBB for subject AA. BBBB can be '' (i.e. mean.nii.gz) which means averaged across all of the NSD core scan sessions; or 'FIRST5' which means averaged across the first 5 NSD core scan sessions (this version was used in various co-registration procedures); or '_sessionNN' which means the Nth NSD core scan session; or '_prffloc' which means the prffloc scan session.
nsddata/ppdata/subjAA/func*/[T1,T2,SWI,TOF]_to_func*.nii.gz
This is a version of the subject’s anatomical volumes that has been matched to the functional data space.
nsddata/ppdata/subjAA/func*/signaldropout.nii.gz
nsddata/ppdata/subjAA/func*/signaldropout_masked.nii.gz
nsddata/freesurfer/subjAA/label/[lh,rh].signaldropout.mgz
These are volumes that indicate areas of EPI signal dropout. They are computed by dividing the T2 (T2_to_func*.nii.gz) volume by the mean pre-processed EPI volume (mean.nii.gz) and then scaling the resulting volume such that 1 corresponds to a reasonable threshold that divides “good” EPI voxels from “bad” ones (see paper for details). The former volume (signaldropout.nii.gz) is not masked, whereas the latter volume (signaldropout.nii.gz) is masked according to the aseg.nii.gz file (any voxel that is zero in aseg is set to zero). The masked volume can be useful for ignoring voxels outside of the brain.
nsddata/freesurfer/subjAA/label/[lh,rh].surfacevoxels*.mgz
Results of the ‘surface voxels’ visualization technique (Kay et al., NeuroImage, 2019). We sampled 1-, 2-, and 3-mm volumetric test patterns onto surface vertices using nearest-neighbor interpolation.
nsddata/ppdata/subjAA/func*/validBBBB.nii.gz
nsddata/freesurfer/subjAA/label/[lh,rh].valid.mgz
This is a binary mask indicating which voxels contain valid data in BBBB for subject AA. (Invalid data occurs when motion or spatial distortion cause missing data for voxels.) BBBB can be '' (i.e. valid.nii.gz) which means average of valid mask across all NSD core scan sessions; or '_sessionNN' which means the Nth NSD core scan session; or '_prffloc' which means the prffloc scan session.
In valid.nii.gz, the values consists of fractions between 0 and 1. For the most part, data were acquired for the entire brain in every session. However, there are a few sessions in which a small amount of brain was cut off. These cases can be detected by finding voxels in valid.nii.gz with values less than 1.0.
