search

faskowit / multiAtlasTT

multi atlas transfer tools for neuroimaging (maTT)
MIT License
57 stars 8 forks source link

AAL2 atlas labels #12

Open 14thibea opened 1 year ago

14thibea commented 1 year ago

Hello @faskowit,

Thanks a lot for this wonderful repo! I already used aicha with freesurfer recon-all and everything went smoothly, and I am now trying to use the aal atlas (corresponding to AAL2?) you put on this repo. However, I am not sure how to link the regions of your version (LH_region_1, LH_region_2...) to the ones of the original atlas (Precentral_L, Frontal_Sup_2_L...). Is there a dictionary somewhere to give an equivalence between them?

faskowit commented 1 year ago

Hi! Thanks for your interest in this repo! For the AAL, I unfortunately do not have a mapping to the original labeling scheme, and it would be a bunch of work to figure out the mapping and to make sure it is correct. Therefore, if the names of the labels are critical, I am afraid that you'll have to use the MNI version of the AAL atlas and find a method to warp the MNI to your individual subjects

Sleosoz commented 2 months ago

Hi, I second the compliments on this repo!! It would indeed be very helpful to count on a mapping to the original region names for the AAL, Arslan, and Power atlases. Is it still not possible to count on those in this repo? Thanks in advance!

faskowit commented 2 months ago

Hi! Thanks again for the compliments and your interest in the tools! But unfortunately, the situation is the same as before... I don't have the bandwidth to assign the regions better, more cognitive/behavioral relevant names. However, I do have a suggestion about how to hack it reasonably, if you would like to do this.

If you fit the AAL labels, for example, with maTT or maTT2 you get a nice cortical ribbon with numbered labels. Then, you could also do a quick and dirty volumetric nonlinear warp using a tool like ANTs, FNIRT, or DARTEL of the Yeo/Scheafer/Yan labels (which have nice, relevant names) in MNI space to the subject space. Then, for each numbered region of the cortical AAL, you could identify the Yeo/Schaefer/Yan node with the greatest voxel overlap in subject space. I think this would be a reasonable way to give names to the regions!