BrainVISA Sulci Atlas (v1)

This dataset provides sulci images and sulcal meshes, for four different atlases. The sulci nomenclature follows the BrainVISA/Morphologist sulci nomenclature [https://brainvisa.info/web/morphologist.html](https://brainvisa.info/web/morphologist.html) version 2018.
- Colin27: MNI public atlas, built from a single subject [https://www.mcgill.ca/bic/software/tools-data-analysis/anatomical-mri/atlases/colin-27](https://www.mcgill.ca/bic/software/tools-data-analysis/anatomical-mri/atlases/colin-27) (1998). The public average T1-weighted MRI image has been segmented using the BrainVISA/Morphologist pipeline and sulci extracted, then they have been manually labelled.
- ICBM152c: MNI public atlas, built from 152 subjects [https://www.bic.mni.mcgill.ca/ServicesAtlases/ICBM152NLin2009](https://www.bic.mni.mcgill.ca/ServicesAtlases/ICBM152NLin2009). The public inter-subject average T1-weighted MRI image has been segmented using the BrainVISA/Morphologist pipeline and sulci extracted, then they have been manually labelled.
- BigBrain: post-mortem brain from Forschungszentrum Jülich [https://julich-brain-atlas.de/atlas/bigbrain](https://julich-brain-atlas.de/atlas/bigbrain) (2015). The public (400μm isotropic resolution) tissue classification image has been imported into the BrainVISA/Morphologist pipeline. Sulci have been automatically extracted, then they have been manually labelled.
- LearnClean: probabilistic atlas built from a dataset of 62 brains, also the training set of BrainVISA/Morphologist sulci model. Each individual subject T1-weighted MRI has been segmented using the BrainVISA/Morphologist pipeline and sulci extracted, then they have been manually labelled. Probabilistic maps of the identified sulci have then been built.

Sulci have been manually identified in all datasets by experts of sulcal anatomy.