FastSurfer stops at mris_sample_parc

[nimzodisaster]

Running docker latest on WSL. Fails to create pial surfaces.

In the recon-surf log a signal 9 terminate signal keeps getting sent:

=========== Creating surfaces rh - map input asegdkt_segfile to surf ===============

mris_sample_parc -ct /opt/freesurfer/average/colortable_desikan_killiany.txt -file /fastsurfer/recon_surf/rh.DKTatlaslookup.txt -projmm 0.6 -f 5 -surf white.preaparc 110021-100_1 rh aparc.DKTatlas+aseg.orig.mgz aparc.DKTatlas.mapped.prefix.annot
Command terminated by signal 9
@#@FSTIME 2023:11:25:21:48:54 mris_sample_parc N 14 e 73.01 S 50.62 U 7.95 P 80% M 12330484 F 1666 R 1972570 W 0 c 298 w 11874 I 42520 O 0 L 2.04 1.98 1.58
@#@FSLOADPOST 2023:11:25:21:50:08 mris_sample_parc N 14 2.08 2.00 1.63

[m-reuter]

Most likely there is already a problem with the input data? You could double check that aparc.DKTatlas+aseg.orig.mgz looks good and also rh.white.preaparc .
freeview norm.mgz aparc.DKTatlas+aseg.orig.mgz -f rh.white.preaparc
(not sure how to run Freeview on WSL, maybe you need to run it in a virtual machine. maybe the FreeSurfer webpages have some details).

[nimzodisaster]

Thanks. Just checked it out. It looks generally okay to me. some pics attached.

[m-reuter]

Yes the white looks ok, but this is not the aparc.DKTatlas+aseg.orig.mgz right? Because that should also have cortical sub-segmentations in it, and here it looks like there is only left and right GM.

[nimzodisaster]

Sorry, I had altered the color scheme. The segmentation looks okay

[nimzodisaster]

build-stamp.txt
deep-seg.log
defect2seg.log
lastcall.build-stamp.txt
patchdir.txt
ponscc.cut.log
recon-all.log
recon-all-status.log
recon-surf.log
unknown-args.txt

all the logs

[m-reuter]

Thanks, unclear why the command mris_sample_parc fails, everything looks good as far as I can see. This happens on a single case or on several cases? Is there a way for you to share this case with us, or is it maybe from a publicly available dataset?

[nimzodisaster]

Thanks. I have emailed you directly @dzne.de

[m-reuter]

I just ran the mris_sample_parccommand from FreeSurfer 7.3.2 directly on your data and it works.
Strange that it fails for you inside the docker (which also uses 7.3.2). I will try a complete run in docker on Linux next.

[m-reuter]

It also runs through with docker. I think it is a memory related issue. We found out that mris_sample_parcrequires much more memory than other FreeSurfer binaries (like >15GB in some cases). Maybe you can increase the memory that is provided to the docker image?

[m-reuter]

I just merged a python replacement for the sample_parc into the dev branch, that should not have these memory issues.