Freesurfer July 2025

freesurfer@nmr.mgh.harvard.edu

42 participants
63 discussions

Re: Freesurfer installation help
by fsbuild 07 Jul '25

07 Jul '25

External Email - Use Caution  Try using the bash shell (by typing "bash" in the terminal window) and setup the environment with that instead of zsh. mymac[~] bash The default interactive shell is now zsh.To update your account to use zsh, please run `chsh -s /bin/zsh`.For more details, please visit https://secure-web.cisco.com/1D7LUoH_cGGJxXRUBJGiZn5rlB8hoE9OQBmiNugxPvMyZ5… bash-3.2$ export FREESURFER_HOME=/Applications/freesurfer/8.0.0bash-3.2$ source /Applications/freesurfer/8.0.0/SetUpFreeSurfer.sh-------- freesurfer-macOS-darwin_x86_64-8.0.0-20250205-0f2bd3a --------Setting up environment for FreeSurfer/FS-FAST (and FSL)FREESURFER_HOME   /Applications/freesurfer/8.0.0FSFAST_HOME       /Applications/freesurfer/8.0.0/fsfastFSF_OUTPUT_FORMAT nii.gzSUBJECTS_DIR      /Applications/freesurfer/8.0.0/subjectsMNI_DIR           /Applications/freesurfer/8.0.0/mni if the above works you could try changing your default shell to bash with "chsh -s /bin/bash" - R.

1 0

Re: Recon-all clinical vs. FS version 8.0.0
by Zollei, Lilla,Ph.D. 07 Jul '25

07 Jul '25

Hi Stephanie, How old are the subjects in your pediatric cohort? Lilla -------- Forwarded Message -------- Subject: [Freesurfer] Recon-all clinical vs. FS version 8.0.0 Date: Tue, 1 Jul 2025 21:48:43 +0000 From: Stefanie Bradley <stefanie.bradley(a)mail.utoronto.ca><mailto:stefanie.bradley@mail.utoronto.ca> Reply-To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu><mailto:freesurfer@nmr.mgh.harvard.edu> To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu><mailto:freesurfer@nmr.mgh.harvard.edu> External Email - Use Caution Hi there, I am analyzing some pediatric brains with clinical pathology, and I was wondering if I would expect a difference in the reconstructions from regular recon-all (Freesurfer version 8.0.0) vs. recon-all clinical, since they both use SynthSeg from what I understand? Thanks, Stefanie

2 1

Ubuntu 24 release
by Alpay Özcan 07 Jul '25

07 Jul '25

External Email - Use Caution Hello, Any plans for compiling on Ubuntu 26.04? Thanks in advance... -- Alpay Özcan, D.Sc. Professor Electrical & Electronics Eng. Dept., Boğaziçi University Bebek Istanbul 34342 Türkiye Tel: +90 212 359 7427 (or 6414 for Dept.)

1 0

mri_vol2vol bug in fs8.0.0?
by Zahnert, Felix 06 Jul '25

06 Jul '25

Dear experts, I ran into an issue running mri_vol2vol after running bbregister using freesurfer 8. Running the same commands in freesurfer 7.4.1 worked well. bbregister worked in both versions, while mri_vol2vol (8.0.0) resulted in the following error: INFO: This REGISTER_DAT transform is valid only for volumes between COR types with c_(r,a,s) = 0. [...] INFO: using NIfTI-1 sform (sform_code=2) error: File exists error: LTAchangeType: src geometry must be valid Running version 7.4.1 mri_vol2vol using the .dat file generated with bbregister version 8.0.0 worked as well. The commands: # bbregister, where fs8 is the freesurfer directory of the subject (SUBJECTS_DIR has been set to 'structural_dir', in which fs8 is located) reg_file="$out_dir/diff2str.dat" bbregister --s fs8 --mov "$b0" --reg "$reg_file" --fslmat "${reg_file%.dat}.mat" --dti # mri_vol2vol cmd below, working in version 7.4.1, but not in version 8.0.0 # with mov being a map of diffusion metrics in the same space as b0 here mri_vol2vol \ --mov "$mov" \ --targ "$structural_dir/fs8/mri/T1.mgz" \ --o "$output" \ --no-save-reg --reg "$reg_file" I also tried the following prior to mri_vol2vol, which did not help: mri_convert -rl "$b0" -rt nearest "$my_mov" "$my_mov_modified" I encountered the same issue when running mri_vol2vol and bbregister with the identical input as --mov. Might there be a bug in version 8's mri_vol2vol, or am I missing something? Thank you for your help! Best, Felix

2 2

Trouble shooting
by andrew cardona 06 Jul '25

06 Jul '25

External Email - Use Caution Good afternoon, I am reaching out as I am trying to trouble shoot one participant, yet it keeps failing and stating the same error "mri_convert: cant determine type of output volume" we have tried changing the file manually to mgz and it works yet when we run recon all it crashes. We also can run mri_info on the file perfectly fine, we also uninstalled and reinstalled freesurfer itself. Any guidance would be greatly appreciated, thank you! *This is what is outputted when running recon all: * Adding options -fix-ento-wm -transfer-base-bfs -fix-vsinus -fix-mca-dura -fix-ga -fix-acj -synthstrip -synthseg -synthmorph fs-check-version --s sub-1009 --o /tmp/tmp.oai3VJ Wed Jul 2 11:36:48 EDT 2025 setenv SUBJECTS_DIR /home/tur31606(a)tu.temple.edu/freesurfer_output cd /home/tur31606(a)tu.temple.edu /usr/local/freesurfer/8.0.0/bin/fs-check-version --s sub-1009 --o /tmp/tmp.oai3VJ -rwxr-xr-x 1 root root 18565 Feb 3 20:30 /usr/local/freesurfer/8.0.0/bin/fs-check-version freesurfer-linux-ubuntu22_x86_64-8.0.0-20250204-0f2bd3a $Id$ Linux CLA19571.tu.temple.edu 6.8.0-60-generic #63~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Apr 22 19:00:15 UTC 2 x86_64 x86_64 x86_64 GNU/Linux pid 451741 Current FS Version freesurfer-linux-ubuntu22_x86_64-8.0.0-20250204-0f2bd3a Subject does not have a bstampfile, copying /usr/local/freesurfer/8.0.0/build-stamp.txt Subject FS Version: freesurfer-linux-ubuntu22_x86_64-8.0.0-20250204-0f2bd3a No constraints on version because REQ=UnSet and FsVerFile=NotThere #@#% fs-check-version match = 1 fs-check-version Done INFO: SUBJECTS_DIR is /home/tur31606(a)tu.temple.edu/freesurfer_output Actual FREESURFER_HOME /usr/local/freesurfer/8.0.0 Linux CLA19571.tu.temple.edu 6.8.0-60-generic #63~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Apr 22 19:00:15 UTC 2 x86_64 x86_64 x86_64 GNU/Linux /home/tur31606(a)tu.temple.edu/freesurfer_output/sub-1009 mri_convert /data/Project_LITE/Imaging/LITE_PAT/bids/sub-1009/ses-01/anat/sub-1009_ses-01_T1w.nii.gz /home/tur31606(a)tu.temple.edu/freesurfer_output/sub-1009/mri/orig/001.mgz mri_convert /data/Project_LITE/Imaging/LITE_PAT/bids/sub-1009/ses-01/anat/sub-1009_ses-01_T1w.nii.gz /home/tur31606(a)tu.temple.edu/freesurfer_output/sub-1009/mri/orig/001.mgz mri_convert: can't determine type of output volume Command exited with non-zero status 1 @#@FSTIME 2025:07:02:11:36:49 mri_convert N 2 e 0.00 S 0.00 U 0.00 P 100% M 6144 F 0 R 270 W 0 c 0 w 2 I 0 O 0 L 0.30 0.56 0.42 @#@FSLOADPOST 2025:07:02:11:36:49 mri_convert N 2 0.30 0.56 0.42 Linux CLA19571.tu.temple.edu 6.8.0-60-generic #63~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Apr 22 19:00:15 UTC 2 x86_64 x86_64 x86_64 GNU/Linux recon-all -s sub-1009 exited with ERRORS at Wed Jul 2 11:36:49 EDT 2025 For more details, see the log file /home/ tur31606(a)tu.temple.edu/freesurfer_output/sub-1009/scripts/recon-all.log To report a problem, see http://secure-web.cisco.com/1lZL5fd1-LIJV20MkpYIehpHZ9KjK6NT_O2y2M8boj9gcNF… Best, Andrew Cardona

2 1

trouble with recon-all FLAIRpial flag
by Mojdeh Zamyadi 06 Jul '25

06 Jul '25

External Email - Use Caution Hi FreeSurfer developers, I'm attempting to use FLAIR images to improve the segmentation quality of some of my subjects after successfully running recon-all -all with the T1 images. After QCing the initial results some segmentation needed improvement which is why I am trying to use the FLAIR images on the recon-all -all results with T1. This is the command I am using: recon-all -s ${subject_id}_T1 -FLAIRpial -autorecon2 -autorecon3 but I am getting errors at the "mris_place_surface" stage. I have also tried manually running mri_normalize first and then running recon-all with the -FLAIRpial flag (command below), but I still get the same errors. mri_normalize $SUBJECTS_DIR/${subject_id}_T1/mri/FLAIR.mgz $SUBJECTS_DIR/${subject_id}_T1/mri/FLAIR.norm.mgz I have attached my recon-all.log file here in case it's of any use. Additional info: 1- recon-all --version freesurfer-macOS-darwin_x86_64-8.0.0-20250205-0f2bd3a 2-Mac info: ProductName: macOS ProductVersion: 15.5 BuildVersion: 24F74 3- uname -a Darwin iMac.local 24.5.0 Darwin Kernel Version 24.5.0: Tue Apr 22 19:53:26 PDT 2025; root:xnu-11417.121.6~2/RELEASE_X86_64 x86_64 Also, a note that we have been able to run recon-all v6 with -FLAIR flag on the same subject on an older Mac system. I would appreciate any insights you might have to how to improve my results using Flair images using v8 of FreeSurfer. Thanks, Mojdeh Zamyadi, MSc (she/her) Research Coordinator (Imaging) Centre for Youth Bipolar Disorder Centre for Addiction and Mental Health 80 Workman Way, Toronto 416-535-8501 ext. 33278 Twitter: @CYBDatCAMH<https://secure-web.cisco.com/1YOInhbSeEqszoMsC5o-oTzaa5eOkFEeEi14Qwx4Faci68…> Website: camh.ca/CYBD<https://secure-web.cisco.com/1-8ab9JMFlspDu1rE_5j9UF0W-Y4o0EXEfZACxBbGkZC9N…> [cid:image001.png@01D3E3D3.C62144B0] If you are in crisis or need urgent help: * Call the Kids Help Phone: 1-800-668-6868 or text 686868 * In Toronto, contact the Youthdale Crisis Service Line: 416-363-9990 * Go to a local hospital emergency room * Call 911 CONFIDENTIALITY NOTICE: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message. AVIS DE CONFIDENTIALIT? : Ce message ?lectronique, ainsi que tout fichier qui y est joint, est reserve ? l'usage exclusif du destinataire vis? et peut contenir des renseignements confidentiels et privil?gi?s. Toute lecture, utilisation, divulgation ou distribution non autoris?e est interdite. Si vous n'?tes pas le destinataire vis?, veuillez en aviser l'exp?diteur par retour de courriel et d?truire toutes les copies du message original.

3 2

Mask-to-eTIV > 1
by Krystalyn Edwards-Calma 06 Jul '25

06 Jul '25

External Email - Use Caution Aloha, I am using Version 7.4.1. After running recon-all for multiple subjects, I noticed the Brain mask is larger than the intracranial volume (Mask-to-eTIV > 1). I attempted multiple different skull stripping methods, however, the mask-to-eTIV is always > 1. Any suggestions on how to fix this? Mahalo for your help. Malama pono, Krystalyn Edwards-Calma

2 3

QC recon- all and manual editing
by Mahta Abbaspour 06 Jul '25

06 Jul '25

External Email - Use Caution Hi, I recently attended your FreeSurfer course in June and found it incredibly helpful—thank you! I’m currently working on a research project on AD patients and have been performing quality control on my FreeSurfer outputs using version 8.0.0. I’d really appreciate your input on a few slices I came across, as I wasn’t sure whether they require manual correction. In this image, do you think the pial surface (red line) near both hippocampus (light blue) are undersegmented? [cid:1f744d31-7e4a-40c3-959e-09280cea965e] This is also another screenshot of the same brain. The pial surface on the left placed inside the WM surface. It's just 1-2 slices but I was wondering what's the best approach to these types of error. [cid:173d0154-9a5e-49f2-a430-c68849f64d80] Thank you for your time. Best

4 3

ASLPrep Recon-all Error
by 岳昊 06 Jul '25

06 Jul '25

External Email - Use Caution Dear FreeSurfer Developers, I am a neuroscience student currently working on analyzing ASL MRI data using ASLPrep. I am encountering difficulties running recon-all during the preprocessing pipeline. Specifically, I am running the command recon-all, and the exact error message displayed is Cmdline: recon-all -autorecon-hemi rh -noparcstats -noparcstats2 -noparcstats3 -nohyporelabel -nobalabels -rh-only -openmp 8 -subjid sub-OAS30001 -sd /fs2/home/bnuhe_user04/250623OASIS_output/sourcedata/freesurfer Stdout: ERROR: Flag -rh-only unrecognized. I encountered an issue similar to one reported on GitHub, suggesting a FreeSurfer version incompatibility. Attempts to resolve this by using FreeSurfer version freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c and freesurfer-linux-centos6_x86_64-7.2.0-20210720-aa8f76b did not resolve the original issue. Upgrading to FreeSurfer version freesurfer-linux-centos7_x86_64-7.4.1-20230613-7eb8460 resulted in a complete failure of ASLPrep. For your reference, the following information may be relevant: 1).FreeSurfer Version: freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c, and freesurfer-linux-centos6_x86_64-7.2.0-20210720-aa8f76b and freesurfer-linux-centos7_x86_64-7.4.1-20230613-7eb8460. 2).Platform: Red Hat Enterprise Linux 8.5 (Ootpa) 3).uname -a: Linux cn77042 4.18.0-348.cn.x86_64 #1 SMP Wed Sep 14 10:27:16 CST 2022 x86_64 x86_64 x86_64 GNU/Linux 4).recon-all.log and the crash file: Please find those files attached to this email. Any guidance you could provide in resolving this issue would be greatly appreciated. Thank you for your time and assistance. Sincerely, Hao Yue

2 1

Recon-all clinical vs. FS version 8.0.0
by Stefanie Bradley 02 Jul '25

02 Jul '25

External Email - Use Caution Hi there, I am analyzing some pediatric brains with clinical pathology, and I was wondering if I would expect a difference in the reconstructions from regular recon-all (Freesurfer version 8.0.0) vs. recon-all clinical, since they both use SynthSeg from what I understand? Thanks, Stefanie

2 4

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer July 2025