Nilearn atlases coordinates and labels

atlas
nilearn

#1

Hello,
I’d like to get the labels and the coordinates of the Craddock_2012()_scorr_mean atlas but didn’t find how to get them.
Also, I can’t find the labels for the basc_mutliscale_2015() and power_2011() atlas.

(I need them to plot my results).

Thanks for your help!


#2

these atlases are generated automatically with clustering, not by manual segmentation so the regions don’t have names --these atlases have no labels.


#3

It is possible to grab the coordinates based on atlases as the input. Please have a look at this example:
https://nilearn.github.io/auto_examples/03_connectivity/plot_atlas_comparison.html#sphx-glr-auto-examples-03-connectivity-plot-atlas-comparison-py
The function find_parcellation_cut_coords does that.

To find labels on BASC. Follow this thread this may give some insights.

For Power atlas, what we return are the MNI coordinates. Following Power et al 2011 paper or Neurosynth can give you the labels of these coordinates ?

Hope this helps.


#4

Thanks for your reply.
My problem now is with the craddock_2012().scorr_mean atlas
If I understand correctly I need to use the function “find_probabilistic_atlas_cut_coords”
Or it returns 43 coordinates
And the parcellation returns 249 regions (i.e., the dimension of my matrices are 249*249 when I’m using the craddock atlas)

Any thoughts?
Thanks


#5

With regards to coordinates extraction on craddock_2012().scorr_mean, using “find_probabilistic_atlas_cut_coords” should return you 43 coordinates.


#6

Right but “NiftiLabelMasker” returns the signal from 249 regions


#7

If you are using coordinates to extract signals, then NiftiSpheresMasker can be used.