these are good questions without good answers. It is hard to compare surface-based vs volume-based. The surface-based may be smaller, but it will only include gray matter whereas a volume-based sphere will be bigger but will include WM and CSF. If you're voxel size is small, then 3mm radius might not get more than one or two of them
On 04/03/2017 09:53 PM, ERIK JAHNER wrote:
What is the appropriate size for a surface based seed?
I am referencing a study with volumetric based seeds with spheres with a radius of 3mm (in fMRI data). I would like to do a similar study using surface based representations. In a volumetric based analysis folding may actually increase the amount of cortical surface represented in a sphere. Using a circle on the surface with a 3 mm radius therefore would not be the same. But…. is the difference so minimal and variable across the cortex that it is not worth considering especially since the seeds are derived from published work and not using a localizer?
Do you know of any published guidelines on generating seed sizes in surface based models, not based on a localizer. Or is this the completely wrong approach? Should I be using the tal coordinates to locate which parcellation of the available atlases that the seed was located in? (for example i have a fusiform face area seed which I have set as a circle 3mm radius)
many thanks for your continued help Erik
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer