Add T2w dorsal rootlets labels by valosekj · Pull Request #158 · spine-generic/data-multi-subject

valosekj · 2024-02-13T19:52:39Z

This PR adds dorsal C2-C8 rootlets labels for T2w images of 24 subjects. The labels were part of the "D5 dataset" used for training and testing of the "M5 model":

21 images were used for the rootlets model training
3 images were used for the evaluation of the inter-rater variability and rootlets model testing. Thus, 5 labels (4 manual raters + STAPLE ground truth, details here) are uploaded for each of these 3 images.

Based on this discussion, I chose the following suffixes according our intranet:

sub-amu01_T2w_label-rootlets_dseg.nii.gz - training label (one label per image)
sub-amu02_T2w_label-rootlets_dseg_desc-rater1.nii.gz - testing label (5 labels per image --> using desc- to distinguish between raters)

The rootlets labels were created from the raw T2w images (i.e., no reorientation or resampling was done) --> I included "SpatialReference": "orig" into JSON sidecars.

@mguaypaq, could you please check the PR and merge? Thanks!

…part of the training dataset D5. For details; see https://github.com/ivadomed/model-spinal-rootlets/tree/main/training

… 4 manual raters and the STAPLE algorithm and were part of the testing dataset D5. For details; see https://github.com/ivadomed/model-spinal-rootlets/tree/main/training and ivadomed/model-spinal-rootlets#17

valosekj · 2024-02-16T16:23:25Z

Hold on with merging.

According to this comment, desc-rater should NOT go at the end as currently done in this PR:

$ tree derivatives/labels/sub-amu02/anat/ 
...
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater1.json
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater1.nii.gz
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater2.json
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater2.nii.gz
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater3.json
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater3.nii.gz
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater4.json
├── sub-amu02_T2w_label-rootlets_dseg_desc-rater4.nii.gz
├── sub-amu02_T2w_label-rootlets_dseg_desc-staple.json
├── sub-amu02_T2w_label-rootlets_dseg_desc-staple.nii.gz
...

jcohenadad · 2024-02-16T16:32:07Z

it would be good if @NathanMolinier can vet this (sorry Nathan, you are now our internal BIDS consultant 😅)

NathanMolinier · 2024-02-16T16:53:08Z

Hold on with merging.

According to this comment, desc-rater should NOT go at the end as currently done in this PR:

Yes indeed, doing this will help us all when trying to identify data for a given dataset. I did not add any specific rules in our intranet to follow a specific order when using entities, but the least we should do is to append the entity label-XXX at the end right before the suffix (which is the last thing we should have before the extension).

valosekj · 2024-02-16T19:28:11Z

but the least we should do is to append the entity label-XXX at the end right before the suffix (which is the last thing we should have before the extension).

Okay! Thanks for the confirmation; I will update the filenames in this branch.

Just a side note:

I did not add any specific rules in our intranet to follow a specific order when using entities

I wrote this function fetching individual BIDS entities using regular expressions based on their keys (e.g., sub-, ses-). The function can be extended to any other entities, and its advantage is that it is "order-agnostic" (thanks to regular expressions). Maybe it might also be useful to you.

NathanMolinier · 2024-02-19T16:18:52Z

I wrote this function fetching individual BIDS entities using regular expressions based on their keys (e.g., sub-, ses-). The function can be extended to any other entities, and its advantage is that it is "order-agnostic" (thanks to regular expressions). Maybe it might also be useful to you.

Great ! That's how BIDS should be used indeed ! Entities are using keys to allow automation ! That's why, for me, BIDS should not use suffixes at all, they're basically just random values not attached to any keys.

valosekj · 2024-03-25T22:10:00Z

I updated the filenames to NOT go with desc-rater at the end in cb54bc7.

For example:

sub-amu02_T2w_label-rootlets_dseg_desc-rater1.nii.gz --> sub-amu02_T2w_desc-rater1_label-rootlets_dseg.nii.gz

@NathanMolinier, could you please confirm, that the rootlets labels are now okay?

NathanMolinier · 2024-03-27T19:45:01Z

Hey @valosekj ! This new naming version is indeed better ! Also, I would like to add a small comment: I see that you are using the suffix _dseg because I imagine you are using discrete segmentations (label values may go higher than 1: not binary). However, ideally, the mapping between label values and the segmented structure should be referenced somewhere (on our intranet like discs and vertebrae or directly within the dataset). So, is this mapping available somewhere ? Apart from that the filenames are good to go !

valosekj · 2024-03-27T20:00:40Z

Hey @valosekj ! This new naming version is indeed better !

Great, thanks for taking the look!

Also, I would like to add a small comment: I see that you are using the suffix _dseg because I imagine you are using discrete segmentations (label values may go higher than 1: not binary).

Exactly! Rootlet labels have values from 2 to 8.

However, ideally, the mapping between label values and the segmented structure should be referenced somewhere (on our intranet like discs and vertebrae or directly within the dataset). So, is this mapping available somewhere ? Apart from that the filenames are good to go !

Good point! Currently, the mapping is avaialble, for example, in the GIF in the rootlets README. But yes, adding a section/tutorial about rootlets to https://spinalcordtoolbox.com is a good idea. I've put it on my to-do list (ivadomed/model-spinal-rootlets#39).

@mguaypaq, can you please merge?

mguaypaq

Looks fine from my end, in terms of git annex get.

mguaypaq · 2024-03-28T22:23:09Z

Ah, the dataset validator is failing because it's too old and doesn't run properly, not because it identified a real issue. I'll have to fix it before merging this PR.

valosekj · 2025-01-31T17:40:55Z

Ah, the dataset validator is failing because it's too old and doesn't run properly, not because it identified a real issue. I'll have to fix it before merging this PR.

Hey @mguaypaq! What's the status of the dataset validator?
We were contacted by a user who wanted to download and use the rootlets labels. (and he couldn't find them under the master branch)

NathanMolinier · 2025-01-31T18:00:44Z

Maybe we should wait before merging this PR we are completely updating the filenames and tree with @sandrinebedard in #176

mguaypaq · 2025-02-03T16:35:42Z

Hey @mguaypaq! What's the status of the dataset validator?

Sorry, I haven't worked on this yet. And sadly, I think there's been negative progress: in December 2024, the bids-validator project switched to a completely new version 2.0, using Deno rather than NodeJS as their runtime, so I have to figure out how to upgrade that too.

valosekj · 2025-03-04T20:25:55Z

Okay, after an in-person discussion with @mguaypaq and @NathanMolinier, I did the following steps to make the rootlets masks from this PR compatible with the latest master (after merging #176):

TL;DR: I reoriented the rootlets labels from LPI to RPI to match the corresponding T2w images.

Bring rootlets labels from the jv/add_rootlets_labels branch to master:

git checkout master
git checkout - -- ./derivatives/labels/sub-mgh01/anat/sub-mgh01_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-ucdavis03/anat/sub-ucdavis03_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-stanford05/anat/sub-stanford05_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-stanford02/anat/sub-stanford02_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona02/anat/sub-barcelona02_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu01/anat/sub-amu01_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-balgrist03/anat/sub-balgrist03_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-balgrist04/anat/sub-balgrist04_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona03/anat/sub-barcelona03_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-balgrist02/anat/sub-balgrist02_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-brnoUhb01/anat/sub-brnoUhb01_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-cmrra02/anat/sub-cmrra02_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-mgh02/anat/sub-mgh02_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-cmrra04/anat/sub-cmrra04_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-cardiff04/anat/sub-cardiff04_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-cardiff02/anat/sub-cardiff02_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater4_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater3_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater2_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-staple_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater1_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater2_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater4_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater3_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater1_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-staple_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-barcelona06/anat/sub-barcelona06_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater4_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater3_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater2_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-staple_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater1_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-amu05/anat/sub-amu05_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-balgrist01/anat/sub-balgrist01_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-balgrist06/anat/sub-balgrist06_T2w_label-rootlets_dseg.nii.gz ./derivatives/labels/sub-geneva01/anat/sub-geneva01_T2w_label-rootlets_dseg.nii.gz

Compare resolution and orientation of the rootlets labels and corresponding T2w images:

for f in $(find . -name "*rootlets*.nii.gz"); do                                                                                                                                                                                                                                          (master|…36)
    label=$f
    image=$(echo "$f" | sed 's|^./derivatives/labels/||; s/_label-rootlets_dseg//')

    echo ""
    echo "$label"
    pixdim "$label"
    orientation "$label"
    echo ""

    # Check if $image contains "desc" and skip if it does
    if [[ $image != *desc* ]]; then
        echo "$image"
        pixdim "$image"
        orientation "$image"
    fi
done

TL;DR: resolution is the same; orientation is different (RPI vs LPI)

details

./derivatives/labels/sub-mgh01/anat/sub-mgh01_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-mgh01/anat/sub-mgh01_T2w.nii.gz
pixdim		[-1.0, 0.800001, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-ucdavis03/anat/sub-ucdavis03_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-ucdavis03/anat/sub-ucdavis03_T2w.nii.gz
pixdim		[-1.0, 0.8, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-stanford05/anat/sub-stanford05_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800036, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-stanford05/anat/sub-stanford05_T2w.nii.gz
pixdim		[-1.0, 0.800036, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-stanford02/anat/sub-stanford02_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799971, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-stanford02/anat/sub-stanford02_T2w.nii.gz
pixdim		[-1.0, 0.799971, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-barcelona02/anat/sub-barcelona02_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-barcelona02/anat/sub-barcelona02_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-amu01/anat/sub-amu01_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800001, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-amu01/anat/sub-amu01_T2w.nii.gz
pixdim		[-1.0, 0.800001, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-balgrist03/anat/sub-balgrist03_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-balgrist03/anat/sub-balgrist03_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-balgrist04/anat/sub-balgrist04_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-balgrist04/anat/sub-balgrist04_T2w.nii.gz
pixdim		[-1.0, 0.8, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-barcelona03/anat/sub-barcelona03_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-barcelona03/anat/sub-barcelona03_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-balgrist02/anat/sub-balgrist02_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-balgrist02/anat/sub-balgrist02_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-brnoUhb01/anat/sub-brnoUhb01_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799976, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]
LPI
sub-brnoUhb01/anat/sub-brnoUhb01_T2w.nii.gz
pixdim		[-1.0, 0.799976, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-cmrra02/anat/sub-cmrra02_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-cmrra02/anat/sub-cmrra02_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-mgh02/anat/sub-mgh02_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-mgh02/anat/sub-mgh02_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-cmrra04/anat/sub-cmrra04_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-cmrra04/anat/sub-cmrra04_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-cardiff04/anat/sub-cardiff04_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-cardiff04/anat/sub-cardiff04_T2w.nii.gz
pixdim		[-1.0, 0.800002, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-cardiff02/anat/sub-cardiff02_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-cardiff02/anat/sub-cardiff02_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater4_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800042, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater3_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800042, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater2_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800042, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-staple_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800042, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-brnoUhb03/anat/sub-brnoUhb03_T2w_desc-rater1_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800042, 0.5, 0.5, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater2_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater4_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater3_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-rater1_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-barcelona01/anat/sub-barcelona01_T2w_desc-staple_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-barcelona06/anat/sub-barcelona06_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-barcelona06/anat/sub-barcelona06_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater4_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800001, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater3_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800001, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater2_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800001, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-staple_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800001, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-amu02/anat/sub-amu02_T2w_desc-rater1_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.800001, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI

./derivatives/labels/sub-amu05/anat/sub-amu05_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-amu05/anat/sub-amu05_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-balgrist01/anat/sub-balgrist01_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-balgrist01/anat/sub-balgrist01_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-balgrist06/anat/sub-balgrist06_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.799999, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-balgrist06/anat/sub-balgrist06_T2w.nii.gz
pixdim		[-1.0, 0.799999, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI
./derivatives/labels/sub-geneva01/anat/sub-geneva01_T2w_label-rootlets_dseg.nii.gz
pixdim		[1.0, 0.8, 0.8, 0.8, 0.0, 0.0, 0.0, 0.0]
LPI
sub-geneva01/anat/sub-geneva01_T2w.nii.gz
pixdim		[-1.0, 0.8, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
RPI

Reorient rootlets labels to RPI to match the corresponding T2w images - 3400331:

for f in $(find . -name "*rootlets*.nii.gz"); do sct_image -i $f -setorient RPI -o $f;done

Push updated rootlets labels following https://github.com/spine-generic/spine-generic/wiki/git-annex#images-niigz

mguaypaq

Everything is fine with git-annex.

valosekj added 2 commits February 13, 2024 11:46

Add rootlets ground truth labels for 21 images. These 21 images were …

74b67aa

…part of the training dataset D5. For details; see https://github.com/ivadomed/model-spinal-rootlets/tree/main/training

Add rootlets ground truth labels for 3 images. These labels come from…

50e011c

… 4 manual raters and the STAPLE algorithm and were part of the testing dataset D5. For details; see https://github.com/ivadomed/model-spinal-rootlets/tree/main/training and ivadomed/model-spinal-rootlets#17

valosekj mentioned this pull request Feb 13, 2024

Push rootlets segmentations ivadomed/model-spinal-rootlets#33

Closed

4 tasks

Do NOT use 'desc-rater' at the filenames end

cb54bc7

valosekj mentioned this pull request Mar 27, 2024

Add a section about rootlets to spinalcordtoolbox.com ivadomed/model-spinal-rootlets#39

Open

Merge branch 'master' into jv/add_rootlets_labels

32fd30b

mguaypaq approved these changes Mar 28, 2024

View reviewed changes

Do NOT use 'desc-staple' at the filenames end

ed16263

valosekj and others added 2 commits March 4, 2025 14:46

Merge branch 'master' into jv/add_rootlets_labels

859c228

reorient rootlets labels to RPI

3400331

mguaypaq approved these changes Mar 4, 2025

View reviewed changes

mguaypaq merged commit 24b83d8 into master Mar 4, 2025
1 check failed

mguaypaq deleted the jv/add_rootlets_labels branch March 4, 2025 21:19

Conversation

valosekj commented Feb 13, 2024

Uh oh!

valosekj commented Feb 16, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jcohenadad commented Feb 16, 2024

Uh oh!

NathanMolinier commented Feb 16, 2024

Uh oh!

valosekj commented Feb 16, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

NathanMolinier commented Feb 19, 2024

Uh oh!

valosekj commented Mar 25, 2024

Uh oh!

NathanMolinier commented Mar 27, 2024

Uh oh!

valosekj commented Mar 27, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mguaypaq left a comment

Choose a reason for hiding this comment

Uh oh!

mguaypaq commented Mar 28, 2024

Uh oh!

valosekj commented Jan 31, 2025

Uh oh!

NathanMolinier commented Jan 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mguaypaq commented Feb 3, 2025

Uh oh!

valosekj commented Mar 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mguaypaq left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

valosekj commented Feb 16, 2024 •

edited

Loading

valosekj commented Feb 16, 2024 •

edited

Loading

valosekj commented Mar 27, 2024 •

edited

Loading

NathanMolinier commented Jan 31, 2025 •

edited

Loading

valosekj commented Mar 4, 2025 •

edited

Loading