Freesurfer

freesurfer@nmr.mgh.harvard.edu

17 participants
30086 discussions

Re: [Freesurfer] recon-all CUDA error
by Chiu, Bryan (PHTH) 08 Jul '14

08 Jul '14

Hi, Thank you for your reply. The CUDA 5.x binaries can be found in the CUDA 5.x legacy drivers for CUDA. As long as you set the PATH=/usr/local/cuda/bin:$PATH and LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH, Freesurfer's cudadetect will look into here for the binaries. I think you should note that cudaDetect (belonging to nVidia) is different than Freesurfer's cudadetect (found in usr/freesurfer/bin). Apologies for the previous misconception. My issue right now is that the GTX 780 (kepler) cant work with 5.0. Do you know if it can work with 5.5? I might try my hand to see if I can compile for a newer version of CUDA. Keep me posted. - Bryan Hi there, I'm currently working on compiling FS source code trying to use CUDA 6, and newer GPUs like Titan series. I haven't got lucky yet, but I will keep on trying... In the meantime I would like to download the CUDA 5 binaries. Are these binaries been published? I can't find anything on the FreeSurfer wiki... In my previous work you mentioned "Improving Execution Performance of FreeSurfer" we used CUDA 3 binaries that I had compiled with --enable-fermi flag. Just to clarify your concern, we never used openmp on the test setup of that study, we have implemented a scheduler to manage K simultaneous recon-all executions. Thank you, 2014-07-02 19:06 GMT+02:00 Chiu, Bryan (PHTH) <bryan.chiu(a)ubc.ca<mailto:bryan.chiu@ubc.ca>>: 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<mailto:bryan.chiu@ubc.ca>> Subject: Re: [Freesurfer] recon-all CUDA error (Chiu, Bryan (PHTH)) To: "freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>" <freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>> Message-ID: <680D9B8D5FEAAF40A1A5878265424CA44A59499D(a)s-itsv-mbx06p.ead.ubc.ca<mailto:680D9B8D5FEAAF40A1A5878265424CA44A59499D@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<mailto:bryan.chiu@ubc.ca>> Subject: [Freesurfer] recon-all CUDA error To: "freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>" <freesurfer(a)nmr.mgh.harvard.edu<mailto:freesurfer@nmr.mgh.harvard.edu>> Message-ID: <680D9B8D5FEAAF40A1A5878265424CA44A594981(a)s-itsv-mbx06p.ead.ubc.ca<mailto:680D9B8D5FEAAF40A1A5878265424CA44A594981@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 _______________________________________________ Freesurfer mailing list Freesurfer(a)nmr.mgh.harvard.edu<mailto:Freesurfer@nmr.mgh.harvard.edu> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail. -- Jordi Delgado Mengual PIC (Port d'Informació Científica)

2 4

Recon-all error
by Emad Ahmadi 08 Jul '14

08 Jul '14

Hello FreeSurfers, Wish you a great summer time! I’m running recon-all on DICOM images of a healthy control, and it's giving me an error. I have attached the log file; I’m running FreeSurfer on a Mac computer and the version I’m using is: freesurfer-Darwin-lion-dev-20140701. I would really appreciate your help. Best, Emad Emad Ahmadi, MD ----------------------------------------------- Postdoc Research Fellow Department of Radiology Massachusetts General Hospital Harvard Medical School 25 New Chardon Street, Suite 400 Boston, MA 02114 Tel: 617 726 5237 Email: emad(a)nmr.mgh.harvard.edu

2 3

Anterior Temporal Lobe / Pial Reconstruction Troubles
by Las Blawimo 08 Jul '14

08 Jul '14

Dear Dr. Fischl, Thanks for your quick reply. Attached you will find some images showing the issues we're having. The images are as follows: 1.) sagittal_1353_norm2+3T_vs_default -> the v5.3 suggested norm2 flags (from the release notes) and the -3T flag vs. the default v5.3 pipeline. Blue is the norm2/3T flag pipeline, yellow is the default. 2.) sagittal_1353_wusl_vs_default ->v5.3 using the -washu_mprage flag, obvious improvement in temporal lobe pial surface. Blue is -washu_mprage flag, yellow is the default v5.3 pipeline 3.) axial_1353_norm2+3T_white -> The white surface, which appears OK using v5.3 with the suggested norm2 parameters and the -3T flag 4.) axial_1353_wusl_white_error -> the white surface, with an obvious error in the lh_postcentral area, using the -washu_mprage flag. Let me know if you have any suggestions on other parameters I should try to improve the temporal lobe pial surface, without compromising the white matter surface reconstruction. Sincerely, Las Message: 5 Date: Mon, 7 Jul 2014 11:14:25 -0400 (EDT) From: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> Subject: Re: [Freesurfer] Anterior Temporal Lobe / Pial Reconstruction Troubles To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <alpine.LRH.2.03.1407071114090.30923(a)nmr.mgh.harvard.edu> Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Hi Las can you send us a couple of images of the problems you are seeing? cheers Bruce On Mon, 7 Jul 2014, Las Blawimo wrote: > Dear FreeSurfer Experts, > > > We are collecting a 0.9mm isotropic FSPGR sequence on a GE MR750 3T scanner. > We're observing, in many subjects, a problem mentioned in previous posts to > the list where the pial surface does not follow the cortex properly in the > anterior temporal lobe region (and also in some frontal regions). The white > matter surface generally looks very good. This is with the default options > from recon-all in v5.3. > > We've also tried running the following options, based on previous posts > (expert options) and the release notes for version 5.2: > > -norm2-b 20 -norm2-n 5 > > and > > -norm2-b 20 -norm2-n 5 -3T > > In general, there is a modest improvement in the medial temporal lobe pial > surface in most subjects. A substantial/desired improvement is observed in > only a few subjects. In general, there was no detrimental effect to either > the white matter or pial surface when using these flags. > > Based on older posts to the list, we've also tried using the following > flags: > > -mprage > > and > > -washu_mprage > > The improvement to the pial surface is much more substantial in almost all > subjects, to some extent with the -mprage flag, mostly with the > -washu_mprage flag. However, this dramatic improvement in pial surface > accuracy was accompanied by many major issues with the white surface in > widespread areas across the brain. Also, some subjects fail to reconstruct > fully due to large defects in the surfaces with this option in place (also > something I've read about happening to people using the -washu otpion on the > list). > > A few questions: > > Would you suggest we also try the options "-norm1-b" and "-norm1-n" in > addition to the norm2 options? If so, any suggested starting values to try? > > Alternatively, would it be sensible to try to use the white surface from the > default pipeline and combine it with (or perhaps initialize) the pial > surface from the -washu_mprage flag? > > Do you have other suggestions for us to try? > > Thank you for your time and help! > > Sincerely, > > Las > >

1 0

mri_convert reads out k_ras from which dicom field
by Fan, Qiuyun 07 Jul '14

07 Jul '14

Dear Freesurfer experts, I used EicomEditor to anonymize the dicom header. When I run mri_convert on the anonymized dicom file, it sets k_ras as NaN. I checked the dcmdump results of both, and I couldn't figure out which fields were erased to cause that problem. I'm attaching the dcmdump info, and mri_convert outputs for more information. Any comments on how to debug this? For example, which field(s) in dicom header does the program read/calculate k_ras from? Thanks a lot! Best, Qiuyun

2 1

talairach fixes - when to do manual intervention
by Elissa McIntosh 07 Jul '14

07 Jul '14

Hello freesurfer experts, I have two questions. First, I've come across some of my subjects who have crooked alignment on the horizontal view, but have seemingly good talairach results. Does it matter that the alignment is crooked as long as talairach is good? My second question is how off does a talairach have to be to merit a manual fix? I've attached an image of someone whose talairach is slightly off making the green line go outside the brain more than it usually does. But I don't know if I'll do more harm by adjusting it slightly. On a related note, does it matter that the green line seems to go outside the brain in the moveable image? Where the green line extends beyond I've had to make manual edits in most of my subjects. Thank you!

1 0

problems with Surface-based Interhemispheric Registration
by angela.favaro＠unipd.it 07 Jul '14

07 Jul '14

Hi all, I have a problem with 'Surface-based Interhemispheric Registration' I followed the instructions and all seem to work, but the command: 'surfreg --s $subject --t fsaverage_sym --lh --xhemi' finish with the error below (indeed in the xhemi/label directory, there is no file) What can I do? Thank you for any help! Angela #@# Cortical Parc rh Mon Jun 30 17:44:52 CEST 2014 /Users/angelafavaro/Desktop/risonanze/studyfree/subj024/xhemi/scripts \n mris_ca_label -l ../label/rh.cortex.label -aseg ../mri/aseg.mgz -seed 1234 subj024/xhemi rh ../surf/rh.sphere.reg /Applications/freesurfer/average/rh.curvature.buckner40.filled.desikan_killiany.2010-03-25.gcs ../label/rh.aparc.annot \n No such file or directory mris_ca_label: could not open label file ../label/rh.cortex.label No such file or directory No such file or directory Darwin host-001.homenet.telecomitalia.it 10.8.0 Darwin Kernel Version 10.8.0: Tue Jun 7 16:33:36 PDT 2011; root:xnu-1504.15.3~1/RELEASE_I386 i386 recon-all -s subj024/xhemi exited with ERRORS at Mon Jun 30 17:44:52 CEST 2014

4 13

neonatal processing stream?
by Alexopoulos, Dimitrios 07 Jul '14

07 Jul '14

Several months ago someone mentioned the possibility of a neonatal processing stream. Is this in development, and if so is there an approximate time frame for release. Thanks Jim The materials in this email are private and may contain Protected Health Information. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying, distribution or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return email.

1 0

mris_divide_parcellation for white matter
by Corinna Bauer 07 Jul '14

07 Jul '14

hello, I was wondering if it was possible to use mris_divide_parcellation on white matter labels. Thanks, Corinna

3 13

longitudinal freesurfer error
by Salil Soman 07 Jul '14

07 Jul '14

Hi, I have followed the steps in the longitudinal qdec tutorial ( https://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalTwoStageModel) but when I try to analyze longitudinal features such as long.thickness-rate I get an error: Error in analyze: Couldn't open /SUBJECTS_DIR/SUBJECT_ID/surf/lh.long.thickness-rate.fwhm10.fsaverage.mgz or .mgh file Any suggestions on what I am doing incorrectly? Best wishes, Sal -- Salil Soman, MD, MS Postdoctoral Research Fellow - Stanford Radiological Sciences Laboratory Fellow - Palo Alto War Related Illness and Injury Study Center WOC Neuroradiology Attending - Veterans Affairs Palo Alto Health Care System

2 1

how to edit extracted masks
by Brianna Damadian 07 Jul '14

07 Jul '14

Hello, I used the mri_label2vol command to extract my aseg.mgz to patient native space (output=aseg-in-rawavg.mgz). I then used the mri_extract_label command as follows to create a mask of the ventricles: mri_extract_label aseg-in-rawavg.mgz 4 5 14 15 43 44 72 ventricles.img When I open this mask on top of the rawavg.mgz volume to look at quality, it seems there are some pieces that are clearly ventricle missing from the mask (see attached). Is there a way to edit this mask or the segmentation to include those areas in the ventricle mask as well as the segmentation volume output for the ventricles? Thanks, Brianna

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