Freesurfer

freesurfer@nmr.mgh.harvard.edu

15 participants
30064 discussions

Re: [Freesurfer] Support for GPU on Freesurfer
by fsbuild 03 Apr '24

03 Apr '24

External Email - Use Caution  Use of parallelization depends upon the order of how the data needs to be processed.  The recon-all command is not  single command, but a wrapper command that runs a sequence or pipeline of many lower level freesurfer commands in a certain order.   It’s not unexpected for recon-all of the subject bert data to take ~8 hours on a consumer (non-server, non-cluster), stand-alone machine, e.g. recent Intel I9 chip with 8 physical (performance) ~3 GHz cores, 8 virtual cores, 32G of DDR4 RAM and an SSD drive.  If you have access to a computing cluster, you could try running recon-all there - which should reduce the processing time. - R. On Apr 3, 2024, at 14:58, Camargo, Aldo <ACamargo(a)som.umaryland.edu> wrote:        External Email - Use Caution        Hi there,I was wondering if there is support for GPU on FreeSurfer. I really would like to use it on GPU because is taking 9 hours per subject to recon-all.Thanks a lot and have a nice day,Aldo CamargoSenior Postdoctoral Fellow at the University of Maryland, Baltimore_______________________________________________Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr..mgh.harvard.edu/mailman/listinfo/freesurfer

1 0

Question about mris_compute_lgi error: /scratch/tp94c480ef_a005_4f9a_85af_ab56b34400cd.load_mgh.m.mgh: Permission denied")
by XinRan Wu 03 Apr '24

03 Apr '24

External Email - Use Caution I am a researcher currently working with the Freesurfer 7.1.1 version for a research project. I have encountered an issue while trying to compute the local gyrification index (LGI). We have pre-existing Freesurfer analysis results, and we want to use the mris_compute_lgi script to calculate the LGI values. However, we are facing difficulties in the first step of running the mris_fill command to generate the ?h.pial.filled.mgz file. After the command completes, we are unable to read the generated ?h.pial.filled.mgz file, and we receive the following error: reading filled volume... /bin/bash: /scratch/tp94c480ef_a005_4f9a_85af_ab56b34400cd.load_mgh.m.mgh: Permission denied ERROR: could not open /scratch/tp94c480ef_a005_4f9a_85af_ab56b34400cd.load_mgh.m.mgh for reading ERROR: loading lh.pial.filled.mgz as MGH Dot indexing is not supported for variables of this type. Error in make_outer_surface (line 25) volume=vol.vol; We have also tried to read the lh.pial.filled.mgz file generated by mris_fill using the load_mgh function individually, but we are still unable to read it successfully. Could you please provide us with some guidance on how to resolve this issue?

1 0

Re: [Freesurfer] [FreeSurfer] Lesion detection at different thresholds
by fsbuild 01 Apr '24

01 Apr '24

External Email - Use Caution  You can try using the Silicon Mac development build on the M2 machine..  The download links is, https://secure-web.cisco.com/1sX8Y1I4GPyZJYzijtfGnav0Xe1MBg-Yv4HzZ07-_su8lv… Right click on the pkg installer and then select the “open with installer.app …” option. - R. On Apr 1, 2024, at 20:40, Martin Juneja <mj70481(a)gmail.com> wrote:        External Email - Use Caution        Hi Koen and Doug,I am trying to do similar analysis but when I run run_samseg command, it gives me error "Illegal Instruction: 4"I am using the FreeSurfer 7.4.1 version on Apple M2 Max MacBook Pro 14.4.1.Is it because run_samseg is incompatible with Apple M2? If so, could you please suggest any solution (if available) how I can run run_samseg on my Apple M2 computer?Thanks.

1 0

recon-all mri_convert error: extra argument ("/usr/local/freesurfer/subjects/maoyipeng/mri/orig/001.mgz")
by 彭翠仪 01 Apr '24

01 Apr '24

External Email - Use Caution Hello FreeSurfer Developers, I'm writing to ask for help while using "recon-all" command. I have converted some dcm format files to to nii format. Then I ran the command "recon-all -s maoyipeng -i first_sT1W_3D_TFE_tra_MR350895.nii -all", an error appeared: mri_convert /home/pengcuiyi/Desktop/QinLab/MRI_human_brain/severe similar/first_sT1W_3D_TFE_tra_MR350895.nii /usr/local/freesurfer/subjects/maoyipeng/mri/orig/001.mgz mri_convert: extra argument ("/usr/local/freesurfer/subjects/maoyipeng/mri/orig/001.mgz") type mri_convert -u for usage Command exited with non-zero status 1 @#@FSTIME 2024:04:01:09:09:04 mri_convert N 3 e 0.00 S 0.00 U 0.00 P 100% M 5972 F 0 R 236 W 0 c 0 w 1 I 0 O 0 L 0.43 0.24 0.19 @#@FSLOADPOST 2024:04:01:09:09:04 mri_convert N 3 0.43 0.24 0.19 Linux ubuntu 5.4.0-150-generic #167~18.04.1-Ubuntu SMP Wed May 24 00:51:42 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux recon-all -s maoyipeng exited with ERRORS at Mon Apr 1 09:09:04 HKT 2024 I've searched the list and didn't find a similar question. I'm a bit new to FreeSurfer, so I am unsure how to resolve this issue. I've attached the recon-all.log if that might help. 1) FreeSurfer version: freesurfer-linux-centos7_x86_64-7.2.0-20210720-aa8f76b 2) uname -a: Linux ubuntu 5.4.0-150-generic #167~18.04.1-Ubuntu SMP Wed May 24 00:51:42 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux 3) recon-all.log: see attached

3 2

Re: [Freesurfer] Freesurfer Waitlist
by Bess Lam 30 Mar '24

30 Mar '24

1 0

mri_volsynth fails but exits with 0 exit status
by Aaron Tanenbaum 29 Mar '24

29 Mar '24

External Email - Use Caution I am currently running FreeSurfer 7.4. I wrote a script using the program mri_volsynth. I noticed for the first time the program was trying to write files where it did not have write permissions. It outputted a few error messages but gave a 0 exit status. It would be nice to have it exit with a nonzero exit status.

2 1

Extracting significant regions - Longitudinal two stage model
by Liliana Wu 29 Mar '24

29 Mar '24

External Email - Use Caution Hi, I ran a longitudinal two stage model analysis in 83 participants comparing two groups. After applying cluster correction, I am seeing several significant regions, and I was wondering if there is a way to extract each significant cluster from each participant. I appreciate any help! Thank you, Liliana

2 1

autorecon1 issue
by Saifullah Tumrani 29 Mar '24

29 Mar '24

External Email - Use Caution I am using freesurfer to preprocess my t1w 3d Mris using autorecon1 which includes first 5 preprocessing techniques till skullstripping it worked well for 4 different datasets including ixi one from open neuro SRBPS but when i am using it for another dataset which has alsmost same configuration it’s talairach transformation fails for some samples and brain matter is removed in the final brainmask.mgz file moreover i applied autorecon-all for this dataset and even the process stopped for some showing error talaraich failed and for some samples the whole pipeline was executed without any error but in manual inspection i found brain matter removed by freesurfer kindly help me solving this problem. I later found bad skull stripping fix at this link FsTutorial/SkullStripFix_freeview - Free Surfer Wiki <https://secure-web.cisco.com/1W6bcIn5Fy404rUlulLdvgZBsqq1rLSJmG1UU4wSK-ttJn…> using some commands but i am facing errors using these command recon-all -skullstrip -wsthresh -clean-bm -no-wsgcaatlas -subjid skullstrip1_before kindly look into this if you have also experience same issues or have any insight on this. A response will be appriciated. Thanks

2 1

Part brain is missing
by Chen Yang 29 Mar '24

29 Mar '24

External Email - Use Caution Dear FreeSurfer Developers, I am trying to do the processing for structural MRI data. I found that part of brain is missing in one data (see Fig.1). Then I tried to change the watershed parameter, but it didnt work at all (no matter which number I use), the brainmask looks the same as before. So I did the manual editing, cloned voxel from T1.mgz to brainmask.mgz. Then I run the -autorecon-pial, but it froze in this step - fix topology. I checked 'wm.mgz', it contained a lot skull and a lot of WM are missing (see Fig.2). I also tried to manually edit the brainmask.mgz and wm.mgz, then run '-autorecon-wm' but it also froze in -fix topology. I put other might relevant information in the end. I've searched in the mail archive but didn't find similar errors. I am not sure if I did anything wrong. Does anyone have any thoughts on how to trouble-shoot this problem? Also, Ive attached the recon-all.log in case it's of any use. Thanks in advance! Best regards, Chen Other might relevant Information: 1) Coil element is wrong. 'HE2,4;NE2'. 2) I got an error message when I wad doing the recon-all first time: watershed error: indices out of bounds. Then I added the command line 'mri_add_xform_to_header $mdir/transforms/talairach.xfm $mdir/nu.mgz $mir/nu.mgz' based on this page <https://secure-web.cisco.com/1uRXBiwoaszcFJpp3kKMMx2gS1DnuVTXjdvnMG63iVHxJH…>. Then I got another error message in recon-all: mri_em_register: could not open brainmask.mgz so I added some cp to nu.mgz. It worked but the part of brain is missing, as I mentioned in the beginning (Fig.1). 1) FreeSurfer version: freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0 2) Platform: Linux 5.10.0-28-amd64 3) uname -a: Linux ramones 5.10.0-28 -amd64 #1 SMP Debian 5.10.209-2 (2024-01-31) x86_64 GNU/Linux 4) recon-all.log: see attached -- Chen Yang PhD Student Independent Research Group 'Milestones of Early Childhood Development' Max Planck Institute for Human Cognitive and Brain Sciences Email:yangc@cbs.mpg.de

2 1

Fw: Discrepancy between mris_preproc and mri_surf2surf results when mapping individual's measures to fsaverage space
by tyzhang＠ibp.ac.cn 29 Mar '24

29 Mar '24

External Email - Use Caution Dear FreeSurfer Experts, Sorry, I think I've found the answer. It's because Jacobian correction was performed for mris_preproc, right? Tongyu Zhang -----原始邮件----- 发件人:tyzhang@ibp.ac.cn 发送时间:2024-03-20 16:31:23 (星期三) 收件人: freesurfer(a)nmr.mgh.harvard.edu 抄送: 主题: Discrepancy between mris_preproc and mri_surf2surf results when mapping individual's measures to fsaverage space Dear FreeSurfer Experts, I'm currently working on mapping cortical morphometric measurements from each individual in the HCP dataset to the fsaverage space. I have noticed that both the mris_preproc and mri_surf2surf commands can achieve this function. However, I found that the vertex-level measurements for cortical thickness, sulcal depth, and curvatures are consistent between these two commands, but the results for surface area and gray matter volume are totally different. I am puzzled by this inconsistency and speculate that perhaps these two commands are calculated on different surfaces? I don't understand why the two commands get different results since the inputs are the same. My command are as follows: (Different surface area values will be obtained in lh.area.fsaverage.mgh and lh.area.fsaverage.gii files) mris_preproc --s id --meas area --target fsaverage --hemi lh –out /Results/CorticalFeatures/id/lh.area.fsaverage.mgh mri_surf2surf --srcsubject id --trgsubject fsaverage --hemi lh --sval /HCP/id/surf/lh.area --tval /Results/CorticalFeatures/id/lh.area.fsaverage.gii Thank you so much! Tongyu Zhang

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer