Manual rootlets labeling to assess inter-rater variability

[valosekj]

To evaluate inter-rater variability in manual rootlets labelling, four raters will manually label T2w images from five subjects.

I randomly picked the following five subjects from the testing dataset D3.
The subjects come from the Spinal Cord Head Position MRI, v1.0.1 dataset and the Spine Generic Multi-subject, r20230223 dataset:

sub-007_ses-headNormal_T2w.nii.gz        # OpenNeuro dataset, headNormal --> neutral position
sub-010_ses-headUp_T2w.nii.gz            # OpenNeuro dataset, headUp --> extension position
sub-amu02_T2w.nii.gz                     # Spine Generic, Siemens Verio scanner
sub-barcelona01_T2w.nii.gz               # Spine Generic, Siemens Prisma-fit scanner
sub-brnoUhb03_T2w.nii.gz                 # Spine Generic, GE Signa-PETMR scanner

The labeling will be level-specific, for example., pixel value 2 for the C2 nerve rootlet, pixel value 3 for the C3 nerve rootlet, etc. We are focused on nerve rootlets C2-T1.

Note: no preprocessing is applied, i.e., labeling is performed on the raw images.
Note: we are labeling only dorsal nerve rootlets; see the tutorial below for details.

A short video tutorial on how to identify the rootlet levels using FSLeyes can be found on this link.

Please use the following suffixes to distinguish between raters:

• label-rootlet_rater1.nii.gz - Theo Mathieu
• label-rootlet_rater2.nii.gz - Raphaëlle Schlienger
• label-rootlet_rater3.nii.gz - Jan Valosek
• label-rootlet_rater4.nii.gz - Olivia Kowalczyk

[valosekj]

Manual rootlets labeling finished; additional details can be found here --> closing.