Resample volume and surface to same orientation

, ,
1

I am trying to resample volume to lh.pial to align fmap and mri surface and have an issues, shown on screenshot.

newplot%20(1)
newplot (1).png700×450 48.6 KB

map is higher than freesurfer surface.

newplot%20(3)
newplot (3).png700×450 29.5 KB

mri_info of both volume and pial are:

fmap volume:

ras xform present
xform info: x_r =   1.0000, y_r =   0.0000, z_r =  -0.0000, c_r =     3.3750
          : x_a =   0.0000, y_a =   1.0000, z_a =  -0.0000, c_a =   -15.7500
          : x_s =   0.0000, y_s =   0.0000, z_s =   1.0000, c_s =    19.5000
Orientation   : RAS

pial surface:

('xras', array([-1.,  0.,  0.])),
('yras', array([ 0.,  0., -1.])),
('zras', array([0., 1., 0.])),
('cras', array([-3.433, 18.419, 28.598]))]))

how resample one of them to the center of coordinates? I suppose i might use xform somehow? Found similar problem in https://github.com/nipy/nibabel/issues/518, @oesteban, Have you solved it?

2

Are you using the hemi-L_space-fsnative_pial.surf.gii files in the fmriprep derivatives? If you’re using the FreeSurfer surface directly, you may not be accounting for the change in origin FreeSurfer makes.

3

I am using lh.pial, the problem solved with adding ‘cras’ to lh.pial coordinates, it could refered to nibabel.freeesurfer.read_geometry loader.

4

Oh, I see. Somehow I read the tags indicating that you were using fMRIPrep outputs, and we provide GIFTI surfaces that have already had that transform applied. Note that if you save the transformed coordinates for later use, you should also zero-out the CRAS fields to keep FreeSurfer tools from applying them again.