1/6: Excited to share final version of our new hierarchical SUBCORTEX ATLAS in @NatureNeuro

27 new functional regions organize across 4 scales and adapt to changing cognitive demands

Full text access: https://rdcu.be/b7N8K 

Well done @yetianmed

@DrBreaky @DanielMargulies

2/6: Scales I-IV of our atlas are freely available in CIFTI & NIFTI formats.

We have also integrated our atlas into existing cortex-only atlases (e.g. Gordon, Glasser) to enable whole-brain connectome mapping:

https://github.com/yetianmed/subcortex

3/6: We used model selection to reconcile two complementary models of brain topography: continuous GRADIENTS and discrete BOUNDARIES.

Boundaries are only drawn if local gradients are sharper than expected under a null model (gray curves in the image).

4/6: Using machine learning techniques, we personalized the group-consensus atlas to account for individual differences in connectivity. All regions were detected in unseen individuals.

5/6: Connectivity gradients varied dynamically in response to changing cognitive demands. For example, the anterior hippocampus became more integrated with the amygdala during task engagement, compared to rest.

6/6: Check out the video by @yetianmed :

Thanks to @gordonneuro for the fantastic commentary: https://www.nature.com/articles/s41593-020-00721-4

Check out the Pursuit article: https://pursuit.unimelb.edu.au/articles/mapping-the-terra-incognita-of-our-brains

Feature in Chinese:
https://mp.weixin.qq.com/s/_0nrHilSllyJbMFdvNttfA

Thanks to our Reviewers & Editor!