Nilearn Dictionary Learning - How to label the resulting regions

Dear nilearn experts,

I have extracted spatially constrained brain regions from resting state fMRI data using nilearns Dictionary Learning algorithm and am now wondering whether there is an efficient way to automatically label the resulting regions. Has anyone done this and could offer some recommendations?

Thanks for your input,

Sebastian

Hello! what kind of labels do you need? If you need only anatomical labels,
maybe the simplest is to compare the components to maps from an atlas (which you
can download with nilearn: see for example
https://nilearn.github.io/modules/generated/nilearn.datasets.fetch_atlas_harvard_oxford.html#nilearn.datasets.fetch_atlas_harvard_oxford).
You can also have a look at https://parietal-inria.github.io/DiFuMo/ which
provides dictionary components with anatomical labels at multiple resolutions
(64, 128, 256, 512, 1024 components).

If you want more diverse labels, you can try to use a meta-analysis tool, such
as NeuroQuery
(https://github.com/neuroquery/neuroquery/blob/master/examples/reverse_search_from_brain_maps_to_terms.ipynb),
or NeuroSynth
(https://github.com/neurosynth/neurosynth/blob/master/examples/decode_images.py)