SurfToSurf Dset interpolation error

Dear AFNI community,

I am having issues with interpolating scalar data between different surface spaces with the SurfToSurf command. For example, taking the SUMA MNI surfaces (, let’s say I want to map the curvature data from icosahedron ld141 space (19,8812 nodes) to the FreeSurfer sphere space (16,8098 nodes). I would use:

SurfToSurf -i_gii std.141.lh.sphere.gii -i_gii lh.sphere.gii -dset std.141.lh.curv.niml.dset -prefix output

Which generates the mesh-to-mesh mapping correctly (output.niml.M2M) but gives me the following error:

Error SUMA_MakeSparseColumnFullSorted:
Node index range is outside [0, N_Node[
Error SUMA_DsetCol2FloatFullSortedColumn:
Failed to get full column vector
– Error SUMA_morphDsetToStd (SUMA_SurfaceToSurface.c:701):
Failed to extract
– Error SurfToSurf (SUMA_SurfToSurf.c:679):
Failed to map std.141.lh.curv.niml.dset

I am fairly new to AFNI/SUMA, so any help would be greatly appreciated!

Kind regards,

For a standard transform, you need to specify the conversion mapping. Consider the following example from the help for MapIcosahedron:

SurfToSurf \
-i_fs std.60.rh.smoothwm.asc \
-i_fs rh.smoothwm.asc \
-prefix std.60. \
-mapfile std.60.rh.niml.M2M \
-dset rh.SOMEDSET.gii.dset

What’s the main reason to map to the Freesurfer sphere space? The standard mesh is already comparable across participants.

Dear Peter,

Thank you for the speedy reply! Taking the example from MapIcosahedron that you highlight, I try and run:

SurfToSurf \
-i_gii std.60.lh.smoothwm.gii \
-i_gii lh.smoothwm.gii \
-prefix output \
-mapfile std.60.MNI_N27_lh.niml.M2M \
-dset std.60.lh.curv.niml.dset

Which works fine, but when I try to run:

SurfToSurf \
-i_gii std.141.lh.smoothwm.gii \
-i_gii lh.smoothwm.gii \
-prefix output \
-mapfile std.141.MNI_N27_lh.niml.M2M \
-dset std.141.lh.curv.niml.dset

I get the same error as above. Note that the only difference is which mesh is the originating one (60 vs 141). In principle the operation should be the same, just mapping from different scales, so I am really at a loss as to what is going wrong.

Regarding the motivation to map to Freesurfer space, that is where I am doing the bulk of my analysis - I received some ROI definitions in Icosahedron Dset format and wish to use them in Freesurfer.