Hi Andrea
what is the background image? This is usually either a failure of the intensity normalization or of the surface deformation/segmentation to adaptively estimate the underlying intensity distributions of the gray/white matter. If you look in recon-all.log for that subject you will find a set of lines like:
setting MIN_GRAY_AT_WHITE_BORDER to 71.1 (was 70) setting MAX_BORDER_WHITE to 112.0 (was 105) setting MIN_BORDER_WHITE to 83.0 (was 85) setting MAX_CSF to 59.1 (was 40) setting MAX_GRAY to 100.0 (was 95) setting MAX_GRAY_AT_CSF_BORDER to 77.0 (was 75) setting MIN_GRAY_AT_CSF_BORDER to 47.2 (was 40)
in your case for example you will probably find that the min_border_white is too high or tha the max_gray is too low. You can set these using the expert opts for both mri_segment and mris_make_surfaces.
cheers Bruce
On Tue, 9 Sep 2014, Andrea Horváth wrote:
Dear Freesurfers,
After running recon-all, the wm surfaces are too close to the pial surface, some gray matter is segmented as white matter as seen in the attached picture (red cross: a bigger cortical part is included in the white matter).
How can I shrink the wm surface?
Thank you for your help!
Andrea