Hello,

I've noticed that the overlays I project onto Surf Ice have more smoothing than I would expect; for example, an activation that lies mainly within the dACC in a volumetric view shows up as extending deep into the corpus callosum in Surf Ice. I see an option in SurfIce under "Preferences" that says "Smooth voxel-based images," but it doesn't seem to affect the overlay if I uncheck it. Is there a way to prevent SurfIce from adding any smoothing?


Best,

-Andy

Hi Chris,

This is a voxelwise dataset (i.e., I am painting a NIFTI dataset onto the template). It is also a cluster corrected image, so only voxels surviving the cluster threshold should be displayed.

-Andy

Andy-

Perhaps you can tell me if your overlay is a voxelwise dataset (e.g. you are painting the surface with a NIfTI *.nii file) or has been converted to a mesh iso-surface (e.g. a OBJ, GII, MZ3, etc mesh). 

If you are loading a NIfTI voxelwise dataset, you adjust the threshold using the two min and max values in the Overlay panel. This is described here
  https://www.nitrc.org/tracker/index.php?func=detail&aid=8092&group_id=984&atid=3730
for example, in the following image the threshold for the reddish overlay has been set to 2..12, setting the range to 1..12 will lead to a larger region seurviving the threshold and a range of 6..12 would lead to a more focal region. Choose you threshold based on your correction for multiple comparisons.
  https://www.nitrc.org/tracker/download.php/984/3730/8092/791/fmri_paint_surface.jpg

On the other hand, when you use the "Advanced/ConvertVoxelwiseVolumeToMesh" you will be asked to select a NIfTI image and a threshold. The size of the resulting blob depends on your threshold. Setting a Value of 1 would result in a larger blog than a stringent threshold like 12. In this latter case, you can use the preferences feature you mentioned to decide what happens at the edge of the brain: since we typically only compute statistics inside the brain there is often a sharp cutoff at the edge of the cortex. This is described here:

https://www.nitrc.org/plugins/mwiki/inde...