Freesurfer

freesurfer@nmr.mgh.harvard.edu

17 participants
30086 discussions

Re: [Freesurfer] recon-all CUDA error
by Chiu, Bryan (PHTH) 02 Jul '14

02 Jul '14

Hi, Turns out if you rename libcudart.so.6.0 to the 5.0 in your respective lib64, you can pass the cudadetect for Freesurfer. However after it will crash with a segfault error. This leads me to believe that I installed correctly. I think Freesurfer only supports up to CUDA 5.0 so far (although I know Freesurfer group stopped supporting the option and left the flags around). Two options exist: Either in a future patch the program can just be enabled to call the newer version of CUDA (or if I can find where it tries to find CUDA from, edit it to accept any version), or to install any version of CUDA less than or equal to 5.0 To others: The OpenMP and -use-gpu flag are quite promising. I took a recon down from 12 hours to 3.5 hours using the -openmp 8 flag for a 4 core processor (i7 4770k). I am currently trying to see if I can speed up faster using the CUDA enabled option. There was a paper titled "Improving the Execution Performance of Freesurfer" which implements a resource scheduler on top of the CUDA and openMP option. No comment on the accuracy of the recon-all using all of these in combination. - Bryan Date: Mon, 30 Jun 2014 16:34:45 +0000 From: "Chiu, Bryan (PHTH)" <bryan.chiu(a)ubc.ca> Subject: Re: [Freesurfer] recon-all CUDA error (Chiu, Bryan (PHTH)) To: "freesurfer(a)nmr.mgh.harvard.edu" <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <680D9B8D5FEAAF40A1A5878265424CA44A59499D(a)s-itsv-mbx06p.ead.ubc.ca> Content-Type: text/plain; charset="iso-8859-1" Hi all, I am still working on this problem. Does anyone know if Freesurfer is asking for CUDA 5.0, except that I have CUDA 6.0 installed? My next steps will be to find out if Freesurfer is calling an older version of CUDA that isn't installed, or trying the CUDA samples. I really want to get this to work. -Bryan Date: Fri, 27 Jun 2014 17:59:37 +0000 From: "Chiu, Bryan (PHTH)" <bryan.chiu(a)ubc.ca> Subject: [Freesurfer] recon-all CUDA error To: "freesurfer(a)nmr.mgh.harvard.edu" <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <680D9B8D5FEAAF40A1A5878265424CA44A594981(a)s-itsv-mbx06p.ead.ubc.ca> Content-Type: text/plain; charset="iso-8859-1" Hi, I'm fairly novice at Linux and I've been trying to setup CUDA for recon-all. I get an error as follows: [TLA@COGMOB-linuxPC ~]$ recon-all -s bert -use-gpu -openmp 8 Testing for CUDA device: /usr/freesurfer/bin/mri_em_register_cuda: error while loading shared libraries: libcudart.so.5.0: cannot open shared object file: No such file or directory Linux COGMOB-linuxPC 2.6.32-431.el6.x86_64 #1 SMP Fri Nov 22 03:15:09 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux recon-all -s bert exited with ERRORS at Fri Jun 27 10:53:29 PDT 2014 For more details, see the log file To report a problem, see http://surfer.nmr.mgh.harvard.edu/fswiki/BugReporting Any idea what is going on? I setup CUDA and my Nvidia drivers according to this tutorial (http://aspratyush.wordpress.com/2012/05/06/install-nvidia-cuda-on-centos-6/) With, of course some slight changes to install the latest toolkit. Thanks in advance, - Bryan

2 1

(no subject)
by Chiu, Bryan (PHTH) 02 Jul '14

02 Jul '14

Hi Chris, If you could get support for CUDA 5.5 it would be marvelous. I'm quite inexperienced with the CUDA programming (consider me an enduser), you would have my full support. I also have a computer here to do any testing if needed. I have a PC with CentOS 6.5, Freesurfer 5.3.0, i7 4770k, 16 Gb RAM, GTX 780.. etc. On that note I have been trying to replicate the methods and the scheduler introduced in "Improving the Execution Performance of Freesurfer" by Delgado et al. . Let me know if you need any help or would like to collaborate. -Bryan Date: Wed, 2 Jul 2014 03:46:01 +0000 From: Chris Adamson <chris.adamson(a)mcri.edu.au> Subject: Re: [Freesurfer] Build with CUDA 5.5? To: Z K <zkaufman(a)nmr.mgh.harvard.edu>, Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <43A4603348D2AC409B1B958EFE5C8B940118273C23(a)MITSMXMBX02V.mcri.edu.au> Content-Type: text/plain; charset="us-ascii" I understand about matching all the different combination of driver versions with linux distros. My plan is to use a centos 6.5 distro and just install the CUDA 5.5 library. I will try and replicate what there is already with CUDA 5.0 with the new library. I will let you know how it goes. I understand about not having resources to devote to a programming task. The CUDA versions are still definitely worth since they are faster than the CPU versions even after openmp and they don't force you to hog all the CPU cores. -----Original Message----- From: Z K [mailto:zkaufman@nmr.mgh.harvard.edu] Sent: Wednesday, 2 July 2014 1:41 AM To: Freesurfer support list; Chris Adamson Subject: Re: [Freesurfer] Build with CUDA 5.5? Chris, This is why we no longer actively support CUDA development in freesurfer and instead have focused on openmp to reduce overall runtime. The complexity of the CUDA programming coupled with the required libraries/driver matching of various builds with various OSes is simply beyond the resources we have to devote to the task. As you know, we currently build against CUDA 5.0, and if your OS supports it, you can run the more computationally intensive tasks (mri_ca_register, etc) using CUDA. Beyond that you are free to attempt a build of the freesurfer source code against the CUDA 5.5 on your Ubuntu system. I can assist with that task by providing the other 3rd party libraries as well as general insight, but beyond that we simply do not have the resources to devout to that task. -Zeke On 06/30/2014 10:19 PM, Chris Adamson wrote: > Guys, > > Would it be possible to get Freesurfer 5.3.0 built with CUDA 5.5? The > newest Ubuntu does not have support for 5.0. > > Thanks in advance, > > Chris Adamson. > > *Dr **Chris Adamson > *Research Officer, Developmental Imaging, *Murdoch Childrens Research > Institute *Murdoch Childrens Research Institute Royal Children's > Hospital Flemington Road, Parkville, Victoria 3052, Australia > www.mcri.edu.au<http://www.mcri.edu.au> <http://www.mcri.edu.au/> > > E chris.adamson(a)mcri.edu.au > T 03 9936 6780 > > > ______________________________________________________________________ > This email has been scanned by the Symantec Email Security.cloud service. > For more information please visit http://www.symanteccloud.com > ______________________________________________________________________ > > > _______________________________________________ > Freesurfer mailing list > Freesurfer(a)nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >

1 0

localGI problems
by elisa veronese 02 Jul '14

02 Jul '14

Dear FS users, I'm trying the -localGI on a linux Centos machine, but I'm facing some problems: 1. should I worry about the *Warning: Unable to open display 'iconic'. You will not be able to display graphics on the screen. *? 2. the script asks me if I wan to remove a file (rm: remove regular file `/tmp/tp726715349117711.load_mgh.8128471.mgh'?). Should I say yes and remove it? Is it important? Should I keep it? I tried answering *no*, but now everything seems to be stuck without performing any subsequent processing. Below there is the output I got. Thank you. Regards, elisa [root@gc02lx fselisa]# recon-all -s 1379341346/ -sd /root/Desktop/fselisa/ -localGI INFO: FreeSurfer build stamps do not match Subject Stamp: freesurfer-Linux-centos4_x86_64-stable-pub-v5.3.0 Current Stamp: freesurfer-Linux-centos6_x86_64-stable-pub-v5.3.0 INFO: SUBJECTS_DIR is /root/Desktop/fselisa Actual FREESURFER_HOME /usr/local/freesurfer -rw-r--r--. 1 root root 1624545 Jun 26 18:08 /root/Desktop/fselisa/1379341346/scripts/recon-all.log Linux gc02lx 2.6.32-358.2.1.el6.x86_64 #1 SMP Tue Mar 12 14:18:09 CDT 2013 x86_64 x86_64 x86_64 GNU/Linux INFO: current FREESURFER_HOME does not match that of previous processing. Current: /usr/local/freesurfer Previous: /home/tmp/ 6004023.clrlcgce03.in2p3.fr/CREAM251988513/DIRAC_lhrUhWpilot/6418014/recon-all.sh-3038867747447129/freesurfer /root/Desktop/fselisa/1379341346/surf #-------------------------------------------- #@# Local Gyrification Index lh Fri Jun 27 09:23:34 CEST 2014 mris_compute_lgi --i lh.pial ================= rm -Rf /root/Desktop/fselisa/1379341346/surf/tmp-mris_compute_lgi-lh.pial ================= ================= mkdir -p /root/Desktop/fselisa/1379341346/surf/tmp-mris_compute_lgi-lh.pial ================= ================= mris_fill -c -r 1 lh.pial /root/Desktop/fselisa/1379341346/surf/tmp-mris_compute_lgi-lh.pial/lh.pial.filled.mgz ================= reading surface from lh.pial... writing filled volume to /root/Desktop/fselisa/1379341346/surf/tmp-mris_compute_lgi-lh.pial/lh.pial.filled.mgz... conforming output volume setting resolution for intermediate calculations to 1.0000 ================= make_outer_surface('/root/Desktop/fselisa/1379341346/surf/tmp-mris_compute_lgi-lh.pial/lh.pial.filled.mgz',15,'/root/Desktop/fselisa/1379341346/surf/tmp-mris_compute_lgi-lh.pial/lh.pial-outer'); exit ================= Warning: Unable to open display 'iconic'. You will not be able to display graphics on the screen. < M A T L A B (R) > Copyright 1984-2012 The MathWorks, Inc. R2012b (8.0.0.783) 64-bit (glnxa64) August 22, 2012 To get started, type one of these: helpwin, helpdesk, or demo. For product information, visit www.mathworks.com. >> reading filled volume... rm: remove regular file `/tmp/tp726715349117711.load_mgh.8128471.mgh'? -- Elisa Veronese - PhD *Research Unit on Brain Imaging and Neuropsychology (RUBIN)* *Inter-University Center for Behavioral Neurosciences (ICBN)University of Udine and University of Verona, ItalyICBN Website: http://icbn.uniud.it/tiki-index.php <http://icbn.uniud.it/tiki-index.php>*

3 4

Re: [Freesurfer] Longitudinal preprocessing troubleshooting
by Chiu, Bryan (PHTH) 02 Jul '14

02 Jul '14

Hi, Attached are 2 screenshots from the same participant at a single timepoint as requested in '4'. Right now I am performing recon-all 'base' on all of the participants with multiple timepoints. I will see if the averaging gets better. Another thing that came up. It turns out one of the participants (who I thought was a brainmask error) had a region of their brain surgically removed (or missing). Should this be excluded, or is your software capable of dealing with this? I will check in with my PI on specific notes regarding this case. - Bryan Date: Tue, 1 Jul 2014 13:51:04 -0400 (EDT) From: Louis Nicholas Vinke <vinke(a)nmr.mgh.harvard.edu> Subject: Re: [Freesurfer] Longitudinal preprocessing troubleshooting To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu> Cc: "Chiu, Bryan $PHTH$" <bryan.chiu(a)ubc.ca> Message-ID: <alpine.LRH.2.03.1407011337470.32500(a)nmr.mgh.harvard.edu> Content-Type: text/plain; charset="utf-8" Hi Bryan, Re: 1a) Did you run the case all the way through recon-all? Even if the skullstripping looks poor, it's possible that the surfaces will turn out ok and might only require some pial edits. Otherwise, this site has some helpful info on rerunning the skullstripping step: https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/SkullStripFix_freeview Re: 1b) If you are referring to voxels located in dura being labelled as cerebral cortex in the aseg then this can be ignored. The cerebral cortex labeled voxels are not used to calculate the morphometry stats. https://surfer.nmr.mgh.harvard.edu/fswiki/MorphometryStats Re: 3) You probably just need to adjust the contrast and brightness, using the window and level sliders on the left in freeview. Re: 4) Could you send a snapshot (or two) of that? -Louis On Thu, 26 Jun 2014, Martin Reuter wrote: > Hi Bryan, > > I can answer the longitudinal question (2), the rest is about cross sectional editing and can probably better > be answered by someone else. > > So yes, I would run the base and check brainmask and surfaces there before going too deep into editing the > cross. > > Best Martin > > > Sent via my smartphone, please excuse brevity. > > -------- Original message -------- > From: "Chiu, Bryan (PHTH)" > Date:06/26/2014 8:06 PM (GMT+01:00) > To: freesurfer(a)nmr.mgh.harvard.edu > Subject: [Freesurfer] Longitudinal preprocessing troubleshooting > > Hi all, > I have a set of data with ~75 subjects with up to 4 timepoints of scans. The structure of the data indicates > that it would be best processed using a longitudinal design. > > I have recon-all most of the scans. There are a few scans that I am worried about and wondering how to treat > each case. I do one sanity check with freeview using a modified version of bert's command:? > "freeview -v $SUBJECTS_DIR/bert/mri/brainmask.mgz -v $SUBJECTS_DIR/bert/mri/aseg.mgz:colormap=lut:opacity=0.2 > -f $SUBJECTS_DIR/bert/surf/lh.white:edgecolor=yellow -f $SUBJECTS_DIR/bert/surf/rh.white:edgecolor=yellow -f > $SUBJECTS_DIR/bert/surf/lh.pial:annot=aparc:edgecolor=red -f > $SUBJECTS_DIR/bert/surf/rh.pial:annot=aparc:edgecolor=red" > > 1a) Skull strip or underanalysis errors: I have attached one screenshot of a very severe case of an > underanalysis. How should I treat this case? I have recently read about Editing a Recon in Freeview, but I'm > not sure how to proceed with this. > > b) Skull strip errors -- Some scans have a very small amount of cerebral cortex labelling into the dura mater/ > skull. Does this warrant the manual removal via freeview of the skull? > > 2) Since I am running this through the longitudinal processing, could I be able to skip #1 hoping that the > averaging of the timepoints into a 'base' will remove these issues? > > 3) I have 2 brains that are completely whited out in Freeview (see attached). However the analysis went through > and it looks good to me. Opening the raw Nifti file in mricroGL does not show this at all. What is going on and > should I be concerned? > > 4) I have 2 other brains that failed the recon-all. One of them is a Talairach fail and the other one failed > mri_watershed. Looking at these brains in mricroGL shows a dark shadow being cast over the brain running > diagonally from the front of the head to the back. I think this is the root of the problem. Is there a way to > fix this or should I consider these scans problematic and exclude them? > > Regards, > > - Bryan > >

1 0

surface to volume visualization
by james pardon 02 Jul '14

02 Jul '14

Dear all, I'm looking for a way to automate surface on volume visualization, to check if the surface reconstruction have been successful on a very large number of subjects. Normally, I would use freeview with -f and -v flags to check and see if it is necessary to edit controls points, etc. As an example, one option to automate this process is to use freeview snapshot option and write a script that runs on several hundred subjects and then show all the resulting images in html format. Running freeview with snapshot option works really good, however, since every freeview run needs X11 there is an overhead here, especially when I'm running on a remote cluster with no X11, which makes running X11 almost impossible. I'm wondering whether any one has a better idea for this? Thanks James

3 5

question about qdec Stats Data Import for Longitudinal Workflow
by Salil Soman 02 Jul '14

02 Jul '14

When following the longitudinal qdec tutorial ( https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel) does using the generate Stats Data Tables option extract data for the difference between time point1 and 2, or something else? Thank you. Salil Soman, MD, MS

4 3

Build with CUDA 5.5?
by Chris Adamson 01 Jul '14

01 Jul '14

Guys, Would it be possible to get Freesurfer 5.3.0 built with CUDA 5.5? The newest Ubuntu does not have support for 5.0. Thanks in advance, Chris Adamson. Dr Chris Adamson Research Officer, Developmental Imaging, Murdoch Childrens Research Institute Murdoch Childrens Research Institute Royal Children's Hospital Flemington Road, Parkville, Victoria 3052, Australia www.mcri.edu.au<http://www.mcri.edu.au/> E chris.adamson(a)mcri.edu.au T 03 9936 6780 ______________________________________________________________________ This email has been scanned by the Symantec Email Security.cloud service. For more information please visit http://www.symanteccloud.com ______________________________________________________________________

2 2

Re: [Freesurfer] question about qdec Stats Data Import for Longitudinal Workflow
by Martin Reuter 01 Jul '14

01 Jul '14

Hi I don't know what you mean. The word 'generate' does not appear on that web page. Best Martin Sent via my smartphone, please excuse brevity. -------- Original message -------- From: Salil Soman <salsoman(a)stanford.edu> Date:07/01/2014 9:50 PM (GMT+01:00) To: free surfer <freesurfer(a)nmr.mgh.harvard.edu> Subject: [Freesurfer] question about qdec Stats Data Import for Longitudinal Workflow When following the longitudinal qdec tutorial (https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel) does using the generate Stats Data Tables option extract data for the difference between time point1 and 2, or something else? Thank you. Salil Soman, MD, MS

1 0

Re: [Freesurfer] longitudinal QDEC question
by Martin Reuter 01 Jul '14

01 Jul '14

Hi salil, the command for that are on the wiki. You'd do a one sample group mean (-osgm) . This asks if the variable of interest is significantly different from zero. Since you ran long mris slopes, that variable would be for example the rate of change or the percent change (one measure per subject). Best Martin Sent via my smartphone, please excuse brevity. -------- Original message -------- From: Salil Soman <salsoman(a)stanford.edu> Date:07/01/2014 9:40 PM (GMT+01:00) To: free surfer <freesurfer(a)nmr.mgh.harvard.edu> Subject: [Freesurfer] longitudinal QDEC question Hi, I have been following the steps for longitudinal processing from here (https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel) and was hoping for some advice on how to setup the GLM. I have a set of data where I have 2 time points for all subjects, have run recon-all, qcache, the base and cross steps of processing. I also ran the long mri_slopes step, created my long.qdec table, then created the simplified table for qdec. I would like to perform a whole brain analysis where I identify any brain regions that changed (volume or cortical thickness) from time point 1 and time point 2. I am a little unclear on what the explanatory variable should be performing this analysis. Any suggestions on how best to structure the analysis would be greatly appreciated. Best wishes, Salil Soman, MD, MS

1 0

longitudinal QDEC question
by Salil Soman 01 Jul '14

01 Jul '14

Hi, I have been following the steps for longitudinal processing from here ( https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel) and was hoping for some advice on how to setup the GLM. I have a set of data where I have 2 time points for all subjects, have run recon-all, qcache, the base and cross steps of processing. I also ran the long mri_slopes step, created my long.qdec table, then created the simplified table for qdec. I would like to perform a whole brain analysis where I identify any brain regions that changed (volume or cortical thickness) from time point 1 and time point 2. I am a little unclear on what the explanatory variable should be performing this analysis. Any suggestions on how best to structure the analysis would be greatly appreciated. Best wishes, Salil Soman, MD, MS

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