We've run into an issue with freesurfer-linux-centos7_x86_64-7.1.0-20200511-813297b (on Rocks release 7.0, kernel 3.10.0-957.12.2.el7.x86_64). It appears some runs are failing at the mris_sphere stage, but not producing any very informative errors, just repeated "singular matrix in quadratic form" at the end of mris_sphere as below. I was wondering if anyone had any suggestions for further troubleshooting please?

recon-all is being invoked in two stages, with a manual brain mask replacing /scratch0/wcoath/2217885.100/sub-11233210_ses-followup/mri/brainmask.mgz in between them. This is being managed by a python script, but I think the sequence being executed is the following:

$ recon-all -i /SAN/medic/insight46/analysis/gradwarp/sub-11233210/ses-followup/anat/sub-11233210_ses-followup_T1w_run-1_desc-gradwarp.nii.gz -T2 /SAN/medic/insight46/analysis/gradwarp/sub-11233210/ses-followup/anat/sub-11233210_ses-followup_T2w_run-1_desc-gradwarp.nii.gz -autorecon1 -noskullstrip -noT2pial -cw256 -cubic -3T -openmp 4 -s sub-11233210_ses-followup -status /SAN/medic/insight46/analysis/freesurfer7/status/sub-11233210_ses-followup_recon_all_status.log -log /SAN/medic/insight46/analysis/freesurfer7/log/sub-11233210_ses-followup_recon_all.log

$ mri_convert /SAN/medic/insight46/analysis/drc_brain/sub-11233210/ses-followup/anat/sub-11233210_ses-followup_T1w_run-1_space-orig_desc-drc-brain-mask.nii.gz /scratch0/wcoath/2217885.100/sub-11233210_ses-followup/mri/drcmask.mgz --cw256 --conform -rt nearest

$ mri_mask -T 1 /scratch0/wcoath/2217885.100/sub-11233210_ses-followup/mri/T1.mgz /scratch0/wcoath/2217885.100/sub-11233210_ses-followup/mri/drcmask.mgz /scratch0/wcoath/2217885.100/sub-11233210_ses-followup/mri/brainmask.mgz

$ recon-all -autorecon2, -autorecon3 -cw256 -cubic -T2pial -openmp 4 -s sub-11233210_ses-followup -status /SAN/medic/insight46/analysis/freesurfer7/status/sub-11233210_ses-followup_recon_all_status.log -log /SAN/medic/insight46/analysis/freesurfer7/log/sub-11233210_ses-followup_recon_all.log

The first recon-all (autorecon1) completes okay, the autorecon2,3 stage fails on occasion (this works for most scans). Its last output looks like this (full copy attached):

@#@FSLOADPOST 2020:06:17:02:37:18 mris_curvature N 12 1.14 1.51 1.48

#--------------------------------------------

#@# Sphere lh Wed Jun 17 02:37:18 BST 2020

/state/partition1/scratch0/wcoath/2217885.100/sub-11233210_ses-followup/scripts

mris_sphere -rusage /state/partition1/scratch0/wcoath/2217885.100/sub-11233210_ses-followup/touch/rusage.mris_sphere.lh.dat -seed 1234 ../surf/lh.inflated ../surf/lh.sphere

setting seed for random number genererator to 1234

7.1.0

== Number of threads available to mris_sphere for OpenMP = 4 ==

reading original vertex positions...

scaling brain by 0.309...

projecting onto sphere...

surface projected - minimizing metric distortion...

MRISunfold() max_passes = 1 -------

tol=5.0e-01, sigma=0.0, host=fry-6, nav=1024, nbrs=2, l_area=1.000, l_dist=1.000

using quadratic fit line minimization

complete_dist_mat 0

rms 0

smooth_averages 0

remove_neg 0

ico_order 0

which_surface 0

target_radius 0.000000

nfields 0

scale 1.000000

desired_rms_height -1.000000

momentum 0.900000

nbhd_size 7

max_nbrs 8

niterations 25

nsurfaces 0

SURFACES 3

flags 0 (0)

use curv 0

no sulc 0

no rigid align 0

mris->nsize 2

mris->hemisphere 0

randomSeed 1234

singular matrix in quadratic form

@#@FSTIME 2020:06:17:02:37:18 mris_sphere N 6 e 256.90 S 2.38 U 723.85 P 282% M 2286104 F 0 R 931286 W 0 c 79152 w 33537 I 0 O 0 L 1.14 1.51 1.48

@#@FSLOADPOST 2020:06:17:02:41:35 mris_sphere N 6 2.62 2.15 1.75

Linux fry-611-142.local 3.10.0-1062.9.1.el7.x86_64 #1 SMP Fri Dec 6 15:49:49 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

recon-all -s sub-11233210_ses-followup exited with ERRORS at Wed Jun 17 02:41:35 BST 2020

For more details, see the log file /SAN/medic/insight46/analysis/freesurfer7/log/sub-11233210_ses-followup_recon_all.log

To report a problem, see http://surfer.nmr.mgh.harvard.edu/fswiki/BugReporting

Thanks for your time,