External Email - Use Caution
Hello FS experts,
I'm trying to reslice a third-party atlas into the same resolution as my
data in MNI space. I am using the following command:
mri_convert
--like T1w_MNI.nii.gz
--resample_type nearest
ICBM152_2009c_AtlasWhite.nii.gz
WhiteAtlas_resampled.nii.gz
Output is:
mri_convert --like T1w_MNI.nii.gz --resample_type nearest
ICBM152_2009c_AtlasWhite.nii.gz WhiteAtlas_resampled.nii.gz
reading from ICBM152_2009c_AtlasWhite.nii.gz...
TR=0.00, TE=0.00, TI=0.00, flip angle=0.00
i_ras = (1, 0, 0)
j_ras = (0, 1, 0)
k_ras = (0, 0, 1)
INFO: transform src into the like-volume: T1w_MNI.nii.gz
writing to WhiteAtlas_resampled.nii.gz...
In the resulting resampled atlas, the boundaries of the ROIs appear to use
an interpolation and not nearest neighbor (i.e., they take values other
than the ROI values in the atlas). Not sure if I can embed images - if so,
see the attached for an example:
[image: image.png]
The output is the same whether I include the --resample_type nearest flag
or leave it out, making me think that mri_convert is ignoring the flag and
using the default interp option.
This can be replicated in the relatively current version
"freesurfer-linux-centos7_x86_64-7.1.0-20200511-813297b" as well as the
older "$Id: recon-all,v 1.379.2.73 2013/05/12 23:15:37 nicks Exp $" (which
we keep around for old pipeline compatibility).
Any ideas on how to fix this so the ROI interfaces take appropriate
(nearest neighbor) values rather than interpolated values?
Thanks all!
Jim H.