Outputs

The application outputs will generally mirror the structure of the inputs. The overall output folder structure can be seen below, with brackets denoting optional terms that are present if the input:

<output_dir>/sub-<label>[/ses-<label>]/anat/

Within this folder, the following (at minimum) can be found:

(a) sub-<label>[_ses-<label>]_desc-RegistrationQCAid.png
(b) sub-<label>[_ses-<label>]_desc-RegistrationQCAid.json
(c) sub-<label>[_ses-<label>]_desc-space-<label>_desc-aseg_dseg.nii.gz
(d) sub-<label>[_ses-<label>]_desc-AsegROIs_scalarstats.tsv
(e) sub-<label>[_ses-<label>]_desc-AsegROIs_scalarstats.json
(f) sub-<label>[_ses-<label>]_space-<label>_desc-<label>_<label>.nii.gz
(g) sub-<label>[_ses-<label>]_space-<label>_desc-<label>_<label>.json

(a,b) Are used to visualize the registration between the quantitative images of interest, and the high resolution anatomical space where the segmentation was originally found. Depending on what relaxometry maps are available, either T2, T1, or PD maps will be used as an underlay with a purple outline showing the outer boundary of the brain (taken from the segmentation image).

(c) Is a copy of the segmentation that has been registered and resampled to the space of the quantitative images using nearest neighbor interpolation.

(d, e) Are created by applying the voxel labels from (c) to any of the (at most) the T1/T2/PD maps in the input qMRI directory. Output statistics include 1st percentile, median, mean, standard deviation, and 99th percentile for each of the region labels found within (c), applied to all available maps. Files with a similar format will also be created if a JSON is passed to the region_groupings_json flag. In this case the “desc” label will contain the name of the JSON (not including the extension), and the contents of the file will be statistics that are calculated for any custom groupings of regions that are specified in the JSON. If a region grouping is identified that does not have any associated voxels in (c), then a n/a value for the associated statistics will be stored within the tsv file. All the statistics within (d) are calculated using the segmentation ROIs that have been resampled into the space of the quantitative maps.

(f, g) Is the result of using the inverse of the transform used for (c) to register any available maps to the space of the input segmentation and high-resolution anatomical from the BIDS input directory. The nifti images following the naming (f) have been resampled using bSpline interpolation. At a minimum there will be one nifti/json pair here, with up to 3 pairs if all T1/T2/PD maps are present.