Hi Rick and Peter,
This crash course in surface based analysis helped alot in clearing up my questions regarding suma tools and integration with Freesurfer.


For a vertex-wise analysis can I supply my native space brain cortical thickness files (after I smooth them with surfsmooth) which have been imported with SUMA_Make_Spec_FS (standardized mesh which matches my MNI template standardized mesh) or do I have to supply my fsaverage warped files for a group analysis?

To get to this point, you only need to run your MNI template through freesurfer a single time. Then run freesurfer on all of your individual subjects. Following that step, run @SUMA_Make_Spec_FS on all of your subjects and on the MNI template. This will automatically create standard meshes (by calling MapIcosahedron for you) for each subject and the MNI template.
Once you have all of your single subject data on standard meshes (either by processing the data directly on standard surfaces or using SurfToSurf to map your data from subject original meshes to the standard mesh).

--Can you clarify what you mean by "processing the data directly on standard surfaces? Do you mean fsaverage files? Have my template and subject brains on a standard surface mesh (Mapsiconahedron)? Or something else?

--Another option you mention is take my original mesh (not processed through Mapsiconahedron/@SUMA_Make_Spec_FS otherwise it would be a "standard mesh") and use SurfToSurf to mach my data to a standard mesh. Is the "standard mesh" you are referring to my MNI template processed through Mapsiconahedron/@SUMA_Make_Spec_FS or the icosahedron?


My plan is to do a surface based analysis of cortical thickness along with whole brain and cortical ribbon VBM analysis. Based on the need to do both of these analyses do you think it will be best to process my native space brain through Freesurfer (so i have original brain space cortical thickness values for vertex-wise analysis) and then 3dQwarp my subject brain to my same Freesurfer processed MNI template. Then run my MNI subject brain through freesurfer again (no need to correct surface defects etc this time around since it should be taken care of already). This way vbm of my cortical ribbon file will be more accurate than the MNI Freesurfer generated files since FS uses affine transforms? Or just apply the subject-to-MNI transform directly to the aseg/cortical ribbon files

The reason I suggested the above plan to run original and MNI warped images through FS is that based on what I have read, it looks like Freesurfer affinely warps brains to MNI space (correct me if I am wrong) and so the MNI warped files from freesurfer will not be as accurate for the VBM analysis as nonlinearly warped brains using 3dQwarp for cortical and subcortical structures. Since I would import these MNI brains through the suma scripts it will automatically run through Mapsiconahedron (similar my MNI template), the brains will have the exact same number of surface nodes as my processed MNI template and display more accurately than using the FreesurferMNI files (affine transform based on fsaverage 305 brain) ?

-Ajay



Edited 1 time(s). Last edit at 12/02/2015 04:09AM by AjaySK.