Bidsification of `ms-mayo-critical-lesions` data

[plbenveniste]

We recently received a new batch of data for the dataset for ms-mayo-critical-lesions since the previous data they sent us wasn't the correct one. The dataset seems completely different from the old one.
What should be done with the data currently in the git-annex dataset:

1. We keep it as it and add it in an exclude file. The new dataset is added (subjects numerotation will start after current max subject index)
2. We remove it completely?
3. We move it to another folder?

What is your opinion @jcohenadad

Cross referencing this issue: sct-pipeline/critical-ms-lesion#1

[jcohenadad]

aoutch, this is annoying. When you say "The dataset seems completely different from the old one", do you mean that:

• the data (ie: DICOM/NIFTI) is the same, but file names are different
  • in this case, we should not duplicate data, and instead work with the newest version (and yes, get rid of the previous one)
• the data are different (ie: not same patients)
  • in this case, we should create another dataset, eg: ms-mayo-critical-lesions-2026
• a bit of both
  • that's the tricky case-- we should identify which are the duplicates, remove the old ones and add the new ones

[plbenveniste]

The data is significantly different and finding an overlap will be pretty hard.

Here is what the preivous datase looked like per subject:

Here is what it looks like now:

I suggest creating a new dataset then.

@mguaypaq could you create the dataset ms-mayo-critical-lesions-2026 please?

[mguaypaq]

Done, and I gave write access to @plbenveniste:
https://data.neuro.polymtl.ca/datasets/ms-mayo-critical-lesions-2026

Do you want me to add ...-2025 to the existing ms-mayo-critical-lesions repo?

[plbenveniste]

Do you want me to add ...-2025 to the existing ms-mayo-critical-lesions repo?

Yes ! I think it makes sense.

[mguaypaq]

Rename done, you may have to update the git remote for your existing clones:
https://data.neuro.polymtl.ca/datasets/ms-mayo-critical-lesions-2025