Dear FreeSurfer experts,

I am using HCP pipelines with my custom modifications of FreeSurferHiResWhite.sh (especially to accomodate full-hires reconstruction) to generate white and pial surfaces.

I tested the v6.0 beta mris_make_surfaces binary to hopefully fix some problems with surfaces I found previously with 5.3.0_HCP version.
This update helped to fix my problems but I found that with v.6.0 there is a new issue with overlaping pial surfaces in hemispherical region.

This issue is not present in 5.3.0 version of mris_make_surfaces.

My command line is

mris_make_surfaces -nsigma_above 2 -nsigma_below 3 -aseg aseg.hires -filled filled.hires -wm wm.hires -mgz -sdir my_subject_dir -orig white.deformed -nowhite -orig_white white.deformed -orig_pial pial -T2 my_subj/mri/T2w_hires.norm  -T1 T1w_hires_norm my_subject_id lh

The previous invocation of mris_make_surfaces in FreeSurferHiResWhite.sh (which does not use T2 refinement):
mris_make_surfaces -variablesigma 4 -white NOWRITE -aseg aseg.hires -orig white.deformed -filled filled.hires -wm wm.hires -sdir my_subject_dir -mgz -T1  T1w_hires_norm my_subject_id lh

produces surfaces which do not overlap.

It seems to me that -aseg option of mris_make_surfaces (preventing surfaces to crossing the midline) is not active in case of -T2 refinement.

I also tried to add -white NOWRITE or -cover_seg parameters, but it did not helped either.

See attached screenshot aseg.hires overlaid with surfaces: blue are surfaces without T2 refinement (without interhemispheric overlap), yellow are surfaces after T2 refinement (with interhemispheric overlap).

Attached is also the log file of my mris_make_surfaces -T2 command.

I also uploaded the subject overlapping_surfs.tar.gz to your FTP file exchange server.

Regards,

Antonin Skoch