Anatomical image preprocessing for mesh generation - usage of T2 image

Discussions
1

Hello everyone!

I’m performing source localization with combined T1 and T2 images.

Here’s my pipeline: I first run FreeSurfer recon-all on the T1 image, import the results into Brainstorm, then register the T2 image to the T1 in SPM using the Reslice the volume option, and finally build an FEM mesh with SimNIBS using both the T1 and T2 images.

My question is: would I obtain better results if I included the T2 image right from the start—running recon-all with T2 and a command such as the one below to refine the pial surface? Any recommendations?

2

Hi @bong516612

If you want just to use FEM mesh in Brainstorm, then you do not need to run freesurfer.
When you call Simnibs from Brainstorm, it will call SPM/CAT for the segmentation of the MRIs [T1 (and T2 is both selected)].

You can simply import your T1 and T2 to Brainstorm and then call SimNibs.

If you want to use Freesurfer, you need to run the segmentation and mesh generation outside Brainstorm, using the Mri2mesh function available within SimNibs [in previous version] .
Please refer to this:

3

Thank you a lot! I totally understand what's going on!

Indeed, I read the article about SimNibs, which says charm can use the output of recon-all and represent better.

Optionally, charm can use the results of FreeSurfer’s recon-all for more accurate representations of smaller sulci in the head meshes (–fs-dir RECONALL_RESULTS_DIR ).

I will test this - but I think they will not different much :slight_smile: