Freesurfer

freesurfer@nmr.mgh.harvard.edu

16 participants
30084 discussions

non-brain tissue segmentation from T1-w images
by Banik, Rana 08 Oct '21

08 Oct '21

External Email - Use Caution Hello All, I am a beginner in freesurfer. Was looking for resources to perform volumetric segmentation(not surface mesh) for non-brain tissues like Muscle-skin, fat from T1-w MR images. Is that feature available in freesurfer? I have used FSL for WM, GM, CSF segmentation. Also looking to protect the inter-subject variability without registering the head volumes to MNI or another atlas. -Rana

2 1

Running -localGI without Matlab
by Soichi Hayashi 07 Oct '21

07 Oct '21

External Email - Use Caution Hello. I am trying to run recon-all with -localGI option. I am running freesurfer on HPC clusters where there is no Matlab license. When I add this option, freesurfer fails with this error message. mris_compute_lgi --i rh.pial ERROR: Matlab is required to run mris_compute_lgi! I am running freesurfer where it has access to MCR runtime to run compiled Matlab script. Is it possible to compile whatever the Matlab script that this module is using so that it can run without the full Matlab installation/license on our HPC systems? Thank you! Soichi

1 0

Re: [Freesurfer] Password
by fsbuild 07 Oct '21

07 Oct '21

External Email - Use Caution Sending the info to your email address shortly. - R. On Oct 6, 2021, at 22:25, Francisco Polito haro <galfox83(a)hotmail.com> wrote:        External Email - Use Caution                External Email - Use Caution        Dear Freesurfer team, Could you provide me the password for the user 'developer' of the VM Freesurfer installation? Many thanks in advance, Best regards,Francisco Polito, LFA_______________________________________________Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://secure-web.cisco.com/1KoZLY6VF10O_YMbs6-9VcK_BEU35SGNNv7OYVZx9BP6Zssr1q5BjsZ8KQVvRnRwRs3gcMn4vNTnu0w9avJv7LT3wL9cEqx6rVgIr4dtoZPvkgY8ClahAzWfN0mbk2Wi2cWgSaSFNM2RaZ8bSxJczJQb_9u-CT_eApU1Ah2qCG5dqh4kB5PAqusBzMzNnsEq8zltg22WFILdiOHxSVQR-xBYr5QQJnl6f5dhHG5ikG5V_UFpi0e2RiuYifD-yytu8AYda7grkiDHDsFDTYSXifg/https%3A%2F%2Fmail.nmr.mgh.harvard.edu%2Fmailman%2Flistinfo%2Ffreesurfer

1 0

Password
by Francisco Polito haro 06 Oct '21

06 Oct '21

External Email - Use Caution Dear Freesurfer team, Could you provide me the password for the user 'developer' of the VM Freesurfer installation? Many thanks in advance, Best regards, Francisco Polito, LFA

1 0

Re: [Freesurfer] Source freesurfer functions into R Studio Terminal
by fsbuild 06 Oct '21

06 Oct '21

External Email - Use Caution Hello Shrikanth, When you source the freesurfer setup script in your shell, e.g., bash, $ source $FREESURFER_HOME/SetUpFreeSurfer.sh - it sets your PATH, SUBJECTS_DIR and a few other variables.  It’s the addition to PATH of something like /usr/local/freesurfer/7.2.0/bin that enables freesurfer commands to be found, e.g.,  $ which mri_convert$ /usr/local/freesurfer/7.2.0/mri_convert So you could cut and paste the absolute path of a freesurfer command to run in the R environment via the system() command, i.e., use whatever is echoed by the “which” command above in the shell where you run freesurfer. You could also try to use a wrapper script to reproduce setting PATH and the freesurfer environment and run that script in the system() call.   Something like (change path to FREESURFER_HOME as need be), $ vi  fs_from_r.sh — start here —#!/usr/bin/env  bash  export FREESURFER_HOME=/usr/local/freesurfer/7.2.0 source $FREESURFER_HOME/SetUpFreeSurfer..sh  # add any freesurfer commands after this — end here — $ chmod 755 fs_from_r.sh Test run it in your bash sell first to see that it works, $ ./fs_from_r.sh Then see if you can run the fs_from_r.sh script via the R system() command.  You probably need to give the absolute path to the script though. - R. On Oct 6, 2021, at 16:13, Yadav, Shrikanth Muralidhar <SMYADAV(a)mgh.harvard.edu> wrote:Hello Freesurfer Developers,I had a freesurfer-Rstudio issue today which I solved by just working on a Linux shell script. However, I wanted to know if it is possible to send freesurfer related commands from R studio out of curiosity.My Problem:I was trying to use freesurfer commands like mri_convert by sending them through the "system()" function in R.   However, I got a "sh: mri_convert command not found" error. This indicates freesurfer is not sourced into the Linux terminal in the R Session. My Question: Is it possible to run freesurfer related commands from the rstudio terminal? What I did:I looked up this problem online and found a similar example where the user needed the function bedtools. But updating the system path still threw the same error with .In addition, I was able to send commands like "dcm2niix" from R using the system function mentioned above.  I am certain this is a sourcing issue.  Would appreciate any insightThank youShrikanth_______________________________________________Freesurfer mailing listFreesurfer@nmr.mgh.harvard.eduhttps://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

1 0

Source freesurfer functions into R Studio Terminal
by Yadav, Shrikanth Muralidhar 06 Oct '21

06 Oct '21

Hello Freesurfer Developers, I had a freesurfer-Rstudio issue today which I solved by just working on a Linux shell script. However, I wanted to know if it is possible to send freesurfer related commands from R studio out of curiosity. My Problem: * I was trying to use freesurfer commands like mri_convert by sending them through the "system()<https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/system>" function in R. * However, I got a "sh: mri_convert command not found" error. This indicates freesurfer is not sourced into the Linux terminal in the R Session. My Question: Is it possible to run freesurfer related commands from the rstudio terminal? What I did: 1. I looked up this problem online and found a similar example<https://stackoverflow.com/questions/26065564/why-cant-i-call-system-functio…> where the user needed the function bedtools. But updating the system path still threw the same error with . 2. In addition, I was able to send commands like "dcm2niix" from R using the system function mentioned above. I am certain this is a sourcing issue. Would appreciate any insight Thank you Shrikanth

2 1

Application portal open for 2022 session of Training in Advanced Statistical Methods in Neuroimaging and Genetics
by ROBERT CARY WELSH 05 Oct '21

05 Oct '21

External Email - Use Caution Hi folks, We are pleased to announce that the application portal is open for the NIH funded 2022 session of the training course. Due to the ongoing COVID-19 pandemic, we have decided to do virtual again this year. The course is an intensive 2-weeks, about 8 hours per day, so please only apply if you know you will be able to commit to coming to the full course. https://secure-web.cisco.com/1z9PgSzXVnC3wHMB7PHwnulIp26my3AIVqOnL0kKgEUEGX… Because of the limited size, attendance is by application only. In general, applicants must be a US citizen or permanent resident, though we do have the ability to accommodate 1 to 3 individuals that do not meet this criteria. Robert Welsh ------------------------------ Robert Welsh, PhD Associate Professor Course Director, Training in Advanced Statistical Methods in Neuroimaging and Genetics Huntsman Mental Health Institute University of Utah robert.c.welsh(a)utah.edu

1 0

Question about network-wise CV
by Martin Juneja 05 Oct '21

05 Oct '21

External Email - Use Caution Hi FreeSurfer experts, We submitted a paper where we used Yeo's 7-network atlas to estimate the hemispheric mean cortical volume of each network using mri_annotation2label --subject X --hemi lh --annotation Yeo2011_7Networks_N1000 --labelbase Left_7Ns The question from the reviewer is: *did the network-wise hemispheric mean CV refer to the mean CV of all brain regions within the network?* Could you please help address this question? Thanks!

1 0

Missed grey matter of sulci
by McKavanagh, Andrea [amckav] 04 Oct '21

04 Oct '21

External Email - Use Caution Hi all, I am having a problem with some post recon-all edits. The white matter in some of my scans has been over estimated in the aparc+aseg file, labelling some of the white matter (e.g by the lingual gyrus) as hypointensities and/or missing out many sulcui entirely (figure 1). However, the wm.mgz file is accurate picking up only the correct white matter in these areas (figure 2). This is happening, on most of the scans I have processed with recon-all. I am using Freesurfer 7.2.0 on Mac BigSur and have ran the same scans through previous version of Freesurfer (6/7.1) with the same problem occurring. I have read on the archives this has previously been a bug in another version, is this the case here? Or is there another correction step I could try? Any suggestions are welcome :) Thanks, BW Andrea [cid:9B03F6B2-0EEC-4035-8303-7364C1852FBF][cid:7C1283D9-0375-4187-A927-0CFB4B6BD5A9]

2 2

Longitudinal analysis: error "no data found in mri/orig"
by Mary DC 04 Oct '21

04 Oct '21

External Email - Use Caution Hello FreeSurfer developers, I have created the "subjects" folder, and I included a large sample of participants inside the folder. I have performed transversal analysis (CROSS), all successfully. I am now performing a longitudinal analysis with the first 8 subjects: Baseline: 001.mgz; 002.mgz;...; 008.mgz Longitudinal: 109.mgz; 110.mgz;...; 116.mgz I followed the steps: https://secure-web.cisco.com/1TZpE_bQz0tJBdgLnT_voDCZ6eYQJTDqKQRLS1v7fMqRfN… 1) BASE analysis was correctly performed for the 8 subjects, and folders (_base) were created 2) LONG analysis: -First subject with baseline data 001.mgz: LONG analysis finished without errors and A01.long.A01_base was created. -First subject with longitudinal data 109.mgz finished with error. -The rest of the subjects finished with error. *Error message in all cases: * *ERROR: no CROSS run data found in /usr/local/freesurfer/subjects/A01_3/mri/orig/. Make sure to have a volume called 001.mgz there.If you have a second run of data call it 002.mgz, etc.See also: http://secure-web.cisco.com/1TxTwmkGwewO59MpGmT0k8qAqls31H-w6Ys9AObOvl31f6H… <http://secure-web.cisco.com/1TxTwmkGwewO59MpGmT0k8qAqls31H-w6Ys9AObOvl31f6H…>Linux VISUAL 5.4.0-81-generic #91~18.04.1-Ubuntu SMP Fri Jul 23 13:36:29 UTC 2021 x86_64 x86_64 x86_64 GNU/Linuxrecon-all -s A01_3.long.A01_base exited with ERRORS at dom oct 3 10:59:53 CEST 2021For more details, see the log file /usr/local/freesurfer/subjects/A01_3.long.A01_base/scripts/recon-all.logTo report a problem, see http://secure-web.cisco.com/1p2mJaCstn9B2xUbzEJL6vTV7W-2WQygHUdN3tVXXu7GPAp… <http://secure-web.cisco.com/1p2mJaCstn9B2xUbzEJL6vTV7W-2WQygHUdN3tVXXu7GPAp…>* However, I think I have correctly named the subjects (001.mgz; 002.mgz; 003.mgz...) and cross-sectional analyses were performed correctly. I can not identify the error. Could you please guide me with this problem? thank you in advance, Best, Mary

1 0

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