Freesurfer

freesurfer@nmr.mgh.harvard.edu

14 participants
30063 discussions

Tkregister2 GUI not responding fully
by Viktorija Pratuseviciute 02 Nov '23

02 Nov '23

External Email - Use Caution Hello, We are trying to process a scan with a failed talairach. We have been able to visualize the misalignment and attempted to use the following command (tkregister2 --s PI163 –fstal) to manual adjust the talairach.xfm based on the tutorial (https://secure-web.cisco.com/1acx6rTd53-Vz7knc_wTrpqifC8PTuiKKanQydzd8QVOyX…) The tkregister2 GUI opens up, however only certain functions work: I can switch between target and moveable, I can also double click on sliding scales to switch coronal, saggital and horizontal views, however I am not able to move any of the sliders including the rotate or translate brain and hence unable to fix the talairach. Is there a way to fix this issue? Or an alternative way to fix the bad talairach output? We are using freesurfer/7.3.1 and have attempted the command on our server command line on both mac and windows with same results. Many thanks, Viki

2 6

Re: [Freesurfer] Slow processing of large defect
by fsbuild 02 Nov '23

02 Nov '23

External Email - Use Caution  If your host machine has 32G of RAM, then it should work to give the VM 16G or perhaps even 24G of RAM. If most all swap space is in use, then it’s likely something is under-running or exhausting the available physical RAM.   Some machines start to swap what’s in memory out to disk files when only 80% of RAM is in use.  But disk I/O from loading swap files in/out of memory can cause the machine to slow down or thrash and never finish processing.  With no consequential or RAM hungry applications running on the host, then the host should be able to essentially run idle with 8G of RAM leaving 24G to assign to the VM. - R. On Oct 3, 2023, at 02:01, Pål Gunnar Larsson <pall(a)ous-hf.no> wrote:        External Email - Use Caution        Thank you for fast reply..The computer has 32Gb. The virtual machine has 8Gb. According to ‘top’ all is used, except 140Mb. Swap has 1Gb of which 780Mb is used. Does swapping use lots of CPU? It is 100% most of the time. Fra: fsbuild <fsbuild@contbay.com> Sendt: 3. oktober 2023 04:33Til: freesurfer@nmr.mgh.harvard.eduKopi: Pål Gunnar Larsson <pall@ous-hf.no>Emne: Re: [Freesurfer] Slow processing of large defect  Not clear if "8G of RAM" below means 8G is assigned to the VM, or the host machine has 8G RAM total.   Assuming the VM has ~8G assigned then the host should have 16-24G RAM total.It reads like the VM is thrashing, e.g., swapping out to disk due to lack of RAM - which can cause a crash.   I would try running the VM on a host machine that can provide more RAM.  So if the VM is currently running with 8G, then try doubling it to 16G (on a host with at least 24 or 32G RAM).- R.On 10/2/23 10:31, Pål Gunnar Larsson (pall(a)ous-hf.no) wrote:        External Email - Use Caution        Hello I am running FS 7.4.1 on Ubuntu 22.10 on Virtualbox 21.10, having 8Gb of ram. I am running a child-MRI where I got a message about a large defect which would run slowly or crash. FS is still running after more than 10 days (70%-100% CPU use), but there are nothing written to disk for 10 days. It is working on retesselation of defect 162.Is there any reason to wait it out? Or will the results be unpredictable? Best Pål G. Larsson Ikke sensitiv _______________________________________________Freesurfer mailing listFreesurfer(a)nmr.mgh.harvard.eduMailScanner has detected a possible fraud attempt from "secure-web.cisco.com" claiming to be https://secure-web.cisco.com/1aeEAtzSDgvMg67gOnpL-fUlNHOFDsR1OYOQuJbl5mIkHQ… note that this e-mail is not secure (encrypted).  If you do not wish to continue communication over unencrypted e-mail, please notify the sender of this message immediately.  Continuing to send or respond to e-mail after receiving this message means you understand and accept this risk and wish to continue to communicate over unencrypted e-mail.  _______________________________________________Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

3 2

Re: [Freesurfer] Using Synthseg segmentations to fix recon-all outputs
by Huang, Yujing 02 Nov '23

02 Nov '23

Hi Maya, I think you can pass extra parameters to mri_synthseg with recon-all option '-expert <opt-file>'. 1. Create file mri_synthseg.opt, and have this line in it mri_synthseg -robust 1. Pass '-expert mri_synthseg.opt' to your recon-all Can you reply to the list so other people can answer it too? Also can you keep previous exchanges? Best, Yujing From: Maya Yablonski <mayay(a)stanford.edu> Sent: Wednesday, November 1, 2023 3:05 PM To: Huang, Yujing <YHUANG43(a)mgh.harvard.edu> Subject: Re: [Freesurfer] Using Synthseg segmentations to fix recon-all outputs External Email - Use Caution Thanks! This seems to be working now. Can I set recon-all to use synthseg with the robust algorithm? I tried the flag -robust as it appears for the command mri_synthseg, but that doesn't seem to be a recognized flag by recon-all. Thanks!

1 1

Finding Intracranial volume
by Mirsol Choi 01 Nov '23

01 Nov '23

External Email - Use Caution Hi fellow freesurfers! I am a Masters student currently working with diffusion images. I would like to ask you experts how I can obtain ICV from my T1/T2 or diffusion images. Is there a certain output from recon-all that produces intracranial volume? Or is there a different command I must use? Thank you all so much for your time. I look forward to hearing from you. Peter

2 2

NIH R25 Course : Training in Advancd Statistical Methods in Neuroimaging and Genetics, Salt Lake City, Utah, April 15-26, 2024. Extended deadline!
by ROBERT CARY WELSH 01 Nov '23

01 Nov '23

External Email - Use Caution Hi everybody, We are pleased to announce that the application portal is still open for the NIH funded 2024 session of the course, “Training in Advanced Statistical Methods in Neuroimaging and Genetics”. We hope you pass this email on to your local neuroimaging community. The course will start on Monday, April 15, 2024 and finish on April 26, 2024 and is to be held in person, in Salt Lake City, Utah, United States at the University Guest House Conference Center. Our course also includes an ongoing (throughout the year) continuing education portion. The course is an intensive 2-weeks, about 8 hours per day of a mix of lectures and hands-on computer labs. neurostatsbootcamp.org<http://secure-web.cisco.com/1GCutdK2bppCWbDssYqZS8SY5rngQjhQqY0KSWKrPj_Czf8…> Due to recently increased funding, we are pleased to announce we can drop our tuition this year to $2250 (from $3000). Strictly financial need scholarships will also still be offered. Tuition includes hotel accommodations at the University Guest House (arrival on April 14 with departure on April 26), breakfast, lunch, day time snacks, three dinners out as a group, and the planned evening activities. All course attendees will arrange for their own transportation to Salt Lake City. The course cost does include a daily pass to the Salt Lake City Trax light rail/commuter system. (Please note that the course instructors' effort is supported by the NIH grant thus avoiding that cost being passed on to the course attendees. This course is strictly not-for-profit.) Because of the limited size and advanced nature of the course, attendance is by application only – please make sure to fully read the website instructions. In general, applicants must be a US citizen, permanent resident, or working with an appointment at a US institution, though we do have the ability to accommodate 3 to 6 individuals that do not meet this criteria. We anticipate the total class size to be between 22 and 25 students. There are opportunities for scholarships which can partially offset hotel accommodations. These are based on true financial need and you will be instructed how to apply if you are admitted to the course. We look forward to seeing you in SLC in 2024! Priority application deadline is Nov 1, 2023; decisions around Dec 5th. Late application deadline is Dec 1, 2023; decisions by Jan 5th, 2024. For more information email us at neurostats(a)g.ucla.edu<mailto:neurostats@g.ucla.edu> or advancedstatisticscourse(a)utah.edu<mailto:advancedstatisticscourse@utah.edu> Robert and Anna ------------------------------ Robert Welsh, PhD Course Director, Training in Advanced Statistical Methods in Neuroimaging and Genetics rcwelsh(a)g.ucla.edu<mailto:rcwelsh@g.ucla.edu> robert.c.welsh(a)utah.edu<mailto:robert.c.welsh@utah.edu> Anna Docherty, PhD Associate Course Director, Training in Advanced Statistical Methods in Neuroimaging and Genetics anna.docherty(a)utah.edu<mailto:anna.docherty@utah.edu> Christina Caldera Course Administrator neurostats(a)g.ucla.edu<mailto:neurostats@g.ucla.edu>

1 0

mri_cvs_register failure
by Antonín Škoch 01 Nov '23

01 Nov '23

External Email - Use Caution Dear experts, I have encountered a total failure of mri_cvs_register --mov my_subject --mni The resulting output file final_CVSmorphed_tocvs_avg35_inMNI152_norm.mgz looks like in the attachment. Could you please give me a hint, where could be a problem, or how to troubleshoot the mri_cvs_register routine? Regards, Antonin Skoch

1 0

Re: [Freesurfer] Using Synthseg segmentations to fix recon-all outputs
by Huang, Yujing 31 Oct '23

31 Oct '23

It looks like seg2recon is also looking for FREESURFER. Set it to your Freesurfer install directory, same as FREESURFER_HOME. From: Maya Yablonski <mayay(a)stanford.edu> Sent: Tuesday, October 31, 2023 7:08 PM To: Huang, Yujing <YHUANG43(a)mgh.harvard.edu> Subject: Re: [Freesurfer] Using Synthseg segmentations to fix recon-all outputs External Email - Use Caution Thanks! That seems to do the trick and it started running, but crashed after a few minutes. The log file shows the last thing that ran was this command: seg2recon --s sub-xx --seg PATH/sub-xx/mri/synthseg.rca.mgz --i /PATH/sub-xx/mri/orig.mgz --threads 1 followed by this error: FREESURFER: Undefined variable. Should I define this variable in addition to FREESURFER_HOME and SUBJECTS_DIR? Where should it point to? Thanks for the help! Maya

1 0

TRACULA Error: Subregion Segmentation aparc+aseg+thalnuc.nii.gz
by Lauren Stephens 31 Oct '23

31 Oct '23

External Email - Use Caution Hi FreeSurfers, I am trying to troubleshoot an error during TRACULA preprocessing and the issue seems related to the creation of aparc+aseg+thalnuc.nii.gz. The subregion segmentation prior to trac-all -prep was successful and mri/ThalamicNuclei.v13.T1.FSvoxelSpace.mgz and other Thalamic Nuclei files were generated. In the archives I saw two others have experienced a similar issue without a resolution in the thread. I am using FreeSurfer 7.3.2 on an iMac Pro with Ventura 13.6. I have attached the log file, log error, and my configuration file. Any help would be greatly appreciated! Thank you, Lauren Stephens Lauren Stephens, MS Assistant Clinical Research Coordinator Department of Psychiatry and Behavioral Science Stanford University School of Medicine E: laurenss(a)stanford.edu P: (650)-725-1716

2 5

Using Synthseg segmentations to fix recon-all outputs
by Maya Yablonski 31 Oct '23

31 Oct '23

External Email - Use Caution Hello, I am working with a pediatric sample for which we are seeing many segmentation errors using the standard recon-all pipeline (version 7.3.2). Running synthseg on the same input files improves the segmentations dramatically, and I would like to be able to fix the surface files (?h.white, ?h.pial) according to the synthseg segmentation. I could not find how to do this in the documentation? Can I place the synthseg.nii.gz file somewhere and rerun recon-all? or autorecon? Would appreciate any pointers to the right way to go about this! Alternatively, I ran recon-all-clinical and that also seems to work well on our data, however the output folder is not identical to the standard recon-all folder, so I can not use that with other preprocessing pipelines that rely on freesurfer output (e.g. fmriprep). I would appreciate any advice on how to leverage the synthseg outputs instead of manually editing the segmentations! Thank you, Maya

3 4

Freesurfer crash when loading "volume" surf
by Mirsol Choi 27 Oct '23

27 Oct '23

External Email - Use Caution Hi there, I have processed a t1 image using recon-all function, and when I try to open the surface file "lh.volume" or "rh.volume", freesurfer crashes everytime. This crash also occurs for other surface files as well, such as "lh.area.pial". Attached is the error message, do you know how I can go about this? For your reference I am using a macOS, running freesurfer 7.4.1. Thank you, I look forward to hearing from you. [image: Screen Shot 2023-10-26 at 6.41.05 PM.png] [image: Screen Shot 2023-10-26 at 6.41.55 PM.png]

2 2

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