Freesurfer

freesurfer@nmr.mgh.harvard.edu

19 participants
30088 discussions

Rejoindre mon réseau sur LinkedIn
by Yuko Yotsumoto via LinkedIn 03 Aug '11

03 Aug '11

LinkedIn ------------ Yuko Yotsumoto souhaite se connecter à vous sur LinkedIn : ------------------------------------------ Je vous invite à faire partie de mon réseau professionnel en ligne sur le site LinkedIn. Accepter l'invitation de Yuko Yotsumoto http://www.linkedin.com/e/-qq80um-gqwbl2ub-2j/qaAr-Z_86TRZ5lXG-dA9kZQ49qSe5… Voir l'invitation de Yuko Yotsumoto http://www.linkedin.com/e/-qq80um-gqwbl2ub-2j/qaAr-Z_86TRZ5lXG-dA9kZQ49qSe5… -- (c) 2011, LinkedIn Corporation

1 0

mri_glmfit
by Mandy Nagy 02 Aug '11

02 Aug '11

One of our volume analyses does not look as it should after we perform the mri_glmfit. Our block design paradigm has 2 separate sessions for each subject with alternating periods of task and rest. We did individual analyses for task-rest for each session. We then concatenated the data into 2 files: one for each session. Then we ran fscalc to add the 2 concatenated files. Up until this point, the data looked like it should. subjects dir: /cluster/manoach/milton/subjects freesurfer 5 However, when we then ran the group analysis using mri_glmfit, we got results that looked abnormal. Our command: mri_glmfit --y mstmct_v_base.nii --glmdir group_analysis/volume/mstmct_v_base/ --osgm We were unsure if we should put in the fsgd file, because we weren't sure which subjects should be included because we added the two sessions using fscalc. (Each subject has two sessions and 2 subject IDS). We realize it could be a problem that arose before the mri_glmfit , but we aren't sure if there was anything else we should consider when running the mri_glmfit itself. Thank you!

2 1

Tracula's sensitivity to recon version
by Shanqing Cai 02 Aug '11

02 Aug '11

Hi, While using traclua, I came across the following problem: it seems to me that the results of tracula are quite sensitive to the version in which the surface reconstruction is done. Please see the attached image. In those examples, both the shape and the size of the same tract of the same subject is different by a lot depending on whether version 5.0.0 or 5.1.0 is used for doing recon-all. As a consequence, the stats generated by Tracula, such as tract mean FA, volume, etc., are also dependent on recon version. I wonder whether this is because I did anything wrong in using tracula? Is this kind of version sensitivity expected? If so, which result should I trust? Thanks a lot! Best regards, Shanqing

2 1

processing OASIS database using freesurfer 5
by Gonzalo Rojas Costa 02 Aug '11

02 Aug '11

Hi: I need to reprocess some images of the OASIS database using freesurfer 5... Should I use only the command "recon-all -s OAS1_0074_MR1 -all", or I must add the option "-expert expert.opts" ?... In the first case freesurfer automatically use thresholds specified in expert.opts ?... Sincerely, Gonzalo Rojas Costa

1 0

Re: [Freesurfer] Poor Intensity or Image Resolution?
by Bruce Fischl 02 Aug '11

02 Aug '11

Hi Antonella Are the sequence parameters the same? The way you convert should have no effect Bruce On Jul 27, 2011, at 11:51 AM, Antonella Kis <atorok9(a)yahoo.com> wrote: > Hi Bruce, > > I highly appreciate your great help/advice. > > Unfortunately, today I have another question: I finished running the recon-all for the new subjects for which I converted the dicom files in the NIFTI format as you advised me, using mri_convert. I was telling you that the .nii.gz file for this new subjects are very big , around 38 MB. Today I was analysing the volumes and I see there is a poor image in comparison with my other brains that were converted in a NIFTI format (having just 4 MB size) by someone else before me. > > My question is: is this poor image due to a different conversion way (I understood the other person used Mricron to convert the files) OR I need to do manually corrections adding control points for the intensity normalization. > > What is your suggestion/advise? > > Many thanks. > Antonella > > From: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> > To: Antonella Kis <atorok9(a)yahoo.com> > Cc: "freesurfer(a)nmr.mgh.harvard.edu" <freesurfer(a)nmr.mgh.harvard.edu> > Sent: Tuesday, July 26, 2011 5:53 PM > Subject: Re: [Freesurfer] Loading atlases > > Hi Antonella > > the atlas is not guaranteed to be correct. If it were we would just take the labels from it :) The only way to tell if it is correct is to look at the segmentations on top of the MR and see if they follow the visible boundary > > To get the surfaces you should do > > tkmedit SUBJECT1 brainmask.mgz lh.white \ > -aux T1.mgz -surfs \ > -segmentation aseg.mgz $FREESURFER_HOME/FreeSurferColorLUT.txt > > cheers > Bruce > > > > On Tue, 26 Jul 2011, Antonella Kis wrote: > > > Hi, > > > > I would like to check if my subcortical segmentation is correct so I was running: > > > > tkmedit SUBJECT1 brainmask.mgz \ > > -aux T1.mgz -surfs \ > > -segmentation aseg.mgz $FREESURFER_HOME/FreeSurferColorLUT.txt > > > > 1). Pial (red line), and white (yellow line) surfaces are shown and I wonder how I can add a green lie for my orig surface? > > 2). After I load the aparc+aseg.mgz is theer a way to load and overlay the original/correct atlas so I can see if my segmentation is correct? > > > > Thank you very much. > > Antonella > > > > > > > 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. > > > <P27.tiff>

4 5

xopts script...
by Gonzalo Rojas Costa 02 Aug '11

02 Aug '11

Hi: In the web page "FSTutorial/TroubleshootingData - Free Surfer Wiki" appears "Try using an xopts script first to change the intensity thresholds being used"... What is a xopts script ?... How can I use a xopts script to change the intensity thresholds ?... Sincerely, Gonzalo Rojas Costa

2 1

Rejoindre mon réseau sur LinkedIn
by laura de giglio via LinkedIn 02 Aug '11

02 Aug '11

LinkedIn ------------ laura de giglio souhaite se connecter à vous sur LinkedIn : ------------------------------------------ Je vous invite à faire partie de mon réseau professionnel en ligne sur le site LinkedIn. Accepter l'invitation de laura de giglio http://www.linkedin.com/e/-qq80um-gquocq3x-16/qaAr-Z_86TRZ5lXG-dA9kZQ49qSe5… Voir l'invitation de laura de giglio http://www.linkedin.com/e/-qq80um-gquocq3x-16/qaAr-Z_86TRZ5lXG-dA9kZQ49qSe5… ------------------------------------------ SAVEZ-VOUS que LinkedIn peut trouver une réponse aux questions les plus complexes ? Publiez ces questions dans la rubrique Réponses de LinkedIn pour profiter de l'expertise des professionnels du monde entier : http://www.linkedin.com/e/-qq80um-gquocq3x-16/ask/inv-23/ -- (c) 2011, LinkedIn Corporation

1 0

Fwd: Problem loading nii
by John Jan Drozd 01 Aug '11

01 Aug '11

But of course, Bruce says the .nii.gz format should have worked, so it does not need to be type mgz format. I would trust Bruce. Sorry for not cc'ing the mailing list earlier. Thanks, John John Drozd Post-Doctoral Fellow Robarts Research Institute University of Western Ontario London, Ontario, Canada http://publish.uwo.ca/~jdrozd2 Begin forwarded message: > From: John Jan Drozd <john.drozd(a)gmail.com> > Date: 1 August, 2011 5:10:07 PM EDT > To: "Tax, C.M.W." <c.m.w.tax(a)student.tue.nl> > Subject: Re: [Freesurfer] Problem loading nii > > Hi, > > I believe the image must be type mgz format for Freesurfer. (For FSL it needs to be type nii or nii.gz.) > > Try typing: > > gunzip t1.nii.gz > mri_convert t1.nii t1.mgz > tkmedit -f t1.mgz > > John > > John Drozd > Post-Doctoral Fellow > Robarts Research Institute > University of Western Ontario > London, Ontario, Canada > http://publish.uwo.ca/~jdrozd2 > > > On 2011-08-01, at 4:55 PM, "Tax, C.M.W." <c.m.w.tax(a)student.tue.nl> wrote: > >> Dear Sir/Madam, >> >> I would like to use freesurfer for cortical flattening, working with the nii format. When I try to load the nii.gz file as described on http://www.fmrib.ox.ac.uk/fsl/freesurfer/index.html, I get the following error: >> >> student@BMT-PC125:~$ tkmedit -f t1.nii.gz >> niiRead(): error opening file ./t1.nii.gz >> >> >> Error: Loading volume t1.nii.gz >> >> Couldn't read the anatomical volume. >> >> Tkmedit couldn't read the volume you specified. >> This could be because the image format wasn't recognized, >> or it couldn't find the proper header, >> or the file(s) were unreadable, >> or it was the wrong size. >> >> >> This image t1.nii.gz was obtained by converting DICOM images with dcm2nii. I also tried to load a example dataset downloaded from http://nifti.nimh.nih.gov/nifti-1/data, but still I get the same error. FSL was able to load both datasets without problems. >> >> Please could you help me with this problem? >> >> Thank you in advance, >> >> Kindest Regards >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer(a)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. >>

1 0

parts of brain not included in segmentation
by Branka Milivojevic 01 Aug '11

01 Aug '11

Hi, I'm a relatively new user of freesurfer, so I don't have much experience ... but I think I've also had problems with recon-all output. Posterior part of the brain has been omitted in a number of subjects. For one of the subjects I have tried to increase the area that is included in the brainmask, but the surfaces were still missing parts of the brain (I don't have output for the original problem). For the same subject, I have then tried to include control points (perhaps too many?), and then I ended up with a really large chunk of the occipital lobe missing.(surfsOnOrig.png & surfsOnBrainmask.png) surfsOnOrig_2.pmg and surfsOnBrainmask_2.png are from another subject, but you can see an example of how much occipital lobe is missing what happens when I run recon-all. (perhaps it is not too much to do a manual edit?) I think the problem might be a large difference in the white/gray contrast between occipipal lobe and the rest of the brain. Could you please tell me if this is fairly normal amount of brain to be missing and if there is anything that I could do to improve the situation. Thanks so much in advance. Cheers, Branka

2 1

Re: [Freesurfer] trac-all -path errors
by Priti Srinivasan 01 Aug '11

01 Aug '11

Hi Ping, This is an issue with the mac build, that we are currently working on fixing. Anastasia, (cc'd) is working on a patch that you can download to resolve this issue. As far as the nb0 value is concerned you're correct, it assumes that the first 7 volumes are b0, but it uses that to compute an average b0 mask, so you should be fine using nb0 = 1. It will then use only the first volume (b0) to create the mask. Please refer to this thread for details on nb0 value in the configuration file: https://mail.nmr.mgh.harvard.edu/pipermail//freesurfer/2011-June/018768.html -Priti > HI Priti, > > I've attached the trac-all.log. The "prior" had not finished, > complaining about memory allocation. > > *** error: can't allocate region > *** set a breakpoint in malloc_error_break to debug > MRIalloc(218, 182, 182): could not allocate 158704 bytes for 14th slice Cannot allocate memory > > > The systems I tested were MAC OS X 10.5.8. with 2 x 2.66 Ghz Quad-Core processors and 16 GB memory, and MAC OS X 10.6 with 20 GB memory. Both of them failed with the same errors. > > In addition, I need to set nb0 = 1 in the configure file in order to run the "prior", although there are 7 b0 sparked in every 7 volumes. I think "nb0=7" means the first 7 volumes will be assumed to > non-diffusion weighted. Correct me if I am wrong. > Thanks. > > Ping > > > > On Fri, Jul 29, 2011 at 11:26 AM, Priti Srinivasan > <rspriti(a)nmr.mgh.harvard.edu> wrote: >> Hi Ping, >> The text files get created in the trac-all -prior stage. Maybe something >> went wrong in that step. Is this the case for all the tracts or just for >> this tract? If you can send us the trac-all.log file under the scripts directory it'll be easier to troubleshoot as to what went wrong. -Priti >>> Hi, >>> I got the error running "trac-all -path -c configure.txt" >>> ERROR: Could not open >>> ${SUBJECTS_DIR}/$Subject/dlabel/diff/lh.cst_AS_avg33_mni_flt_cpts_5.txt I can only find only image files in the "dlabel/diff" directory, but not text files. >>> Which step(s) create the lh.cst_AS_avg33_mni_flt_cpts_5.txt file? Thanks, >>> Ping >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer(a)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. >

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