Currently, you'll have to run recon-all to the end in order to run the hippo subfield seg

On 5/30/2022 5:55 AM, Jacob Levenstein wrote:

        External Email - Use Caution

Hello Freesurfer team + users,

I am looking to run segmentHA_T1.sh following samseg + samseg2recon, and came across this previous thread where it seemed to workout:

MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg66074.html

Given that I am writing in, I have not yet gotten this to run properly...

Below is my samseg command:

samseg --t1w ${subject}_${ses}_T1w.nii.gz --sd . --s ${subject}_${ses}_samseg --refmode t1w --threads 16 --parallel

From the log file, samseg seems to run correctly,  but once it reaches the samseg2recon steps, I receive the following error:

ln -sf orig.mgz rawavg.mgz

ERROR: ln -sf orig.mgz rawavg.mgz

ERROR: samseg2recon --s ${subject}_${ses}_samseg --tmpdir

${SUBJECTS_DIR}/${subject}_${ses}_samseg/mri/samseg/tmp.pyBLVF/tmp.mergeseg --cleanup

If I then (re)run samseg2recon on the samseg directory (see cmd below) , I can execute the failed lines, and this script completes without error.

   samseg2recon --s ${subject}_${ses}_samseg

However, when I then try to run: segmentHA_T1.sh ${subject}_${ses}_samseg, I receive the following error:

Cannot find wmparc.mgz or norm.mgz or talairach.xfm for the subject.

Has the subject been procesed with recon-all?

As the error message correctly states, within that subject's dir, I do not have an mri/norm.mgz or mri/wmparc.mgz. I do have an mri/samseg/transforms/samseg.talairach.xfm, but it is not in the scripts expected location/naming convention.

If it is helpful to know:

i. Running recon-all then segmentHA_T1.sh completes without error...

ii. Copy over the mri dir from the recon-all to the samseg subject dir, then running segmentHA_T1.sh also completes without error...

iii. I am running Ubuntu 20.04.4, freesurfer-linux-ubunu18x86_64_-7.2.0-20210721

Any idea what may be going wrong with samseg/samseg2recon? I am happy to use the workaround listed above, but I suspect there may be a more efficient way, and or unintended implications when mixing and matching recon-all and samseg...?

Thank you for your help and insight!

All the best,

Jacob
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
MailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer