Freesurfer July 2013

freesurfer@nmr.mgh.harvard.edu

165 participants
332 discussions

Re: [Freesurfer] motion dependent resolution: bbr vs flt
by Anastasia Yendiki 16 Jul '13

16 Jul '13

Is the motion in the anatomicals bad enough that it would affect the quality of the surfaces? Bbregister uses the surfaces, flirt does not. The "eddy current correction" is also motion correction - it aligns all the frames to the first with an affine registration. Rotation and translation due to head motion, in addition to warping due to eddy currents, is what cause the frames to be misaligned. It's assumed that by aligning them you correct for both (approximately). On Tue, 16 Jul 2013, Pfuhl, Gerit wrote: > Hi, > > we have more motion in the anatomical data between the groups (healthy controls vs patients). > The difference between 5.1 using flt and 5.2. using bbr is not extreme. > If I understood correctly, since we use first recon-all and thatis independent of flt or bbr, any difference musts come how the two algorithms deal with the dti raw data. Is there any motion correction? > The preprocessing steps include eddy current correction but I did not find a step for motion correction. > > Regards > Gerit > ________________________________________ > Von: Anastasia Yendiki [ayendiki(a)nmr.mgh.harvard.edu] > Gesendet: Dienstag, 16. Juli 2013 20:03 > An: Roschinski, Benjamin > Cc: freesurfer(a)nmr.mgh.harvard.edu; Pfuhl, Gerit; King, Joseph > Betreff: Re: [Freesurfer] motion dependent resolution: bbr vs flt > > Hi Benjamin - You can see how bbregister works in this paper: > http://surfer.nmr.mgh.harvard.edu/pub/articles/greve.2009.ni.63.BBR.pdf > > In short it uses the surfaces from the freesurfer recon (which of course > come from the anatomical) to align the anatomical to an EPI (in this case > the low-b from the diffusion series). Because bbregister uses the extra > information about the surfaces (which flirt doesn't), that's why Doug was > asking if the motion is in the anatomical or the diffusion. > > Between 5.1 and 5.2 several things changed in tracula, only one of them > being the default registration method. You can still switch back to flirt > with 5.2 though if you want to check if that's what's making the > difference. The instructions for this are under "Specify the intra-subject > registration method" here: > http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/Tracula > > Let me know if you have any other questions. > > a.y > > On Tue, 16 Jul 2013, Roschinski, Benjamin wrote: > >> Hi Doug, >> >> thanks for the answer. >> We did a comparison between the results of TRACULA 5.1 and 5.2. >> When we compared the two different results (5.1 and 5.2) from health controls and recovered anroxia nervosa patients for the first time the results of TRACULA 5.2 seemed to be better than the ones of TRACULA 5.1 >> Then we did the same with our acute anorexia patients which are less settled in mri studies. >> In the end the TRACULA 5.2 outputs were not as good as the results of TRACULA 5.1 after we included the acute patients. >> I hope now you understand why we thought that bbr might be more sensitive for motion. >> If bbr is pretty robust in anatomical and dti what do you think about flt. Is it as good as bbr or might be better in case of motion? >> Another question: we know that bbregister needs to have a high resolution image reference for its algorithm. >> Is that the T1? >> >> thank you very much >> >> Benjamin >> >> ________________________________________ >> Von: freesurfer-bounces(a)nmr.mgh.harvard.edu [freesurfer-bounces(a)nmr.mgh.harvard.edu]" im Auftrag von "Douglas N Greve [greve(a)nmr.mgh.harvard.edu] >> Gesendet: Montag, 15. Juli 2013 19:41 >> An: freesurfer(a)nmr.mgh.harvard.edu >> Betreff: Re: [Freesurfer] motion dependent resolution: bbr vs flt >> >> Hi Benjamin, BBR does not need higher resolution. Do you mean motion in >> the anatomical or motion in the DTI? BBR should be pretty robust in >> either case. It is hard to know what might be going on without more >> details of why you think it is getting worse. >> doug >> >> >> On 07/14/2013 08:01 AM, Roschinski, Benjamin wrote: >>> Dear freesurfer experts and users, >>> >>> we visualize white matter tracts of anorexia nervosa patients which >>> are characterized by more movements in mri studies. >>> We have the impression that we get worser TRACULA outputs by using >>> bbregister than flt. >>> Could it be possible that bbregister needs a higher resolution than flt? >>> So summarizing it is more sensitive for motion? >>> >>> kind regards >>> >>> Benjamin >>> >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer(a)nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> -- >> Douglas N. Greve, Ph.D. >> MGH-NMR Center >> greve(a)nmr.mgh.harvard.edu >> Phone Number: 617-724-2358 >> Fax: 617-726-7422 >> >> Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting >> FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 >> www.nmr.mgh.harvard.edu/facility/filedrop/index.html >> Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ >> >> _______________________________________________ >> 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. >> >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer(a)nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> > > >

1 0

Re: [Freesurfer] motion dependent resolution: bbr vs flt
by Gerit Pfuhl 16 Jul '13

16 Jul '13

> > Hi, > > we have more motion in the anatomical data between the groups (healthy > controls vs patients). > The difference between 5.1 using flt and 5.2. using bbr is not extreme. > If I understood correctly, since we use first recon-all and thatis > independent of flt or bbr, any difference musts come how the two algorithms > deal with the dti raw data. Is there any motion correction? > The preprocessing steps include eddy current correction but I did not find > a step for motion correction. > > Regards > Gerit > ________________________________________ > Von: Anastasia Yendiki [ayendiki(a)nmr.mgh.harvard.edu] > Gesendet: Dienstag, 16. Juli 2013 20:03 > An: Roschinski, Benjamin > Cc: freesurfer(a)nmr.mgh.harvard.edu; Pfuhl, Gerit; King, Joseph > Betreff: Re: [Freesurfer] motion dependent resolution: bbr vs flt > > Hi Benjamin - You can see how bbregister works in this paper: > > http://surfer.nmr.mgh.harvard.edu/pub/articles/greve.2009.ni.63.BBR.pdf > > In short it uses the surfaces from the freesurfer recon (which of course > come from the anatomical) to align the anatomical to an EPI (in this case > the low-b from the diffusion series). Because bbregister uses the extra > information about the surfaces (which flirt doesn't), that's why Doug was > asking if the motion is in the anatomical or the diffusion. > > Between 5.1 and 5.2 several things changed in tracula, only one of them > being the default registration method. You can still switch back to flirt > with 5.2 though if you want to check if that's what's making the > difference. The instructions for this are under "Specify the intra-subject > registration method" here: > http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/Tracula > > Let me know if you have any other questions. > > a.y > > On Tue, 16 Jul 2013, Roschinski, Benjamin wrote: > > > Hi Doug, > > > > thanks for the answer. > > We did a comparison between the results of TRACULA 5.1 and 5.2. > > When we compared the two different results (5.1 and 5.2) from health > controls and recovered anroxia nervosa patients for the first time the > results of TRACULA 5.2 seemed to be better than the ones of TRACULA 5.1 > > Then we did the same with our acute anorexia patients which are less > settled in mri studies. > > In the end the TRACULA 5.2 outputs were not as good as the results of > TRACULA 5.1 after we included the acute patients. > > I hope now you understand why we thought that bbr might be more > sensitive for motion. > > If bbr is pretty robust in anatomical and dti what do you think about > flt. Is it as good as bbr or might be better in case of motion? > > Another question: we know that bbregister needs to have a high > resolution image reference for its algorithm. > > Is that the T1? > > > > thank you very much > > > > Benjamin > > > > ________________________________________ > > Von: freesurfer-bounces(a)nmr.mgh.harvard.edu [ > freesurfer-bounces(a)nmr.mgh.harvard.edu]" im Auftrag von > "Douglas N Greve [greve(a)nmr.mgh.harvard.edu] > > Gesendet: Montag, 15. Juli 2013 19:41 > > An: freesurfer(a)nmr.mgh.harvard.edu > > Betreff: Re: [Freesurfer] motion dependent resolution: bbr vs flt > > > > Hi Benjamin, BBR does not need higher resolution. Do you mean motion in > > the anatomical or motion in the DTI? BBR should be pretty robust in > > either case. It is hard to know what might be going on without more > > details of why you think it is getting worse. > > doug > > > > > > On 07/14/2013 08:01 AM, Roschinski, Benjamin wrote: > >> Dear freesurfer experts and users, > >> > >> we visualize white matter tracts of anorexia nervosa patients which > >> are characterized by more movements in mri studies. > >> We have the impression that we get worser TRACULA outputs by using > >> bbregister than flt. > >> Could it be possible that bbregister needs a higher resolution than flt? > >> So summarizing it is more sensitive for motion? > >> > >> kind regards > >> > >> Benjamin > >> > >> > >> _______________________________________________ > >> Freesurfer mailing list > >> Freesurfer(a)nmr.mgh.harvard.edu > >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > > > -- > > Douglas N. Greve, Ph.D. > > MGH-NMR Center > > greve(a)nmr.mgh.harvard.edu > > Phone Number: 617-724-2358 > > Fax: 617-726-7422 > > > > Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting > > FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 > > www.nmr.mgh.harvard.edu/facility/filedrop/index.html > > Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ > > > > _______________________________________________ > > 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. > > > > > > _______________________________________________ > > Freesurfer mailing list > > Freesurfer(a)nmr.mgh.harvard.edu > > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > > > > > >

1 0

Freesurfer faster with CUDA or I7 3770k ?
by Dorian P. 16 Jul '13

16 Jul '13

Dear list, We are buying a computer dedicated mostly to Freesurfer. Initially we thought an I7 3770k a bit overclocked (about 4.5ghz) should be a good choice. However, I noticed that Freesurfer seem better tuned to use CUDA. Many of the messages I read are from 2010 so I am not sure whether this is the case. In sum is it better to focus on CUDA gpu or on Intel I7? Second, does the CUDA speeds matter a lot? Third, is it true that I will not be able to run multiple instances of Freesurfer with CUDA, which is possible with I7? Will it still be faster using CUDA rather than running a Freesurfer instance on each I7 core? Fourth, is the OS important, for example can we just continue on Ubuntu or is it better Debian/CentOS etc.? Thank you for your help and apologies if the questions are too basic. Dorian TJU

2 1

motion dependent resolution: bbr vs flt
by Roschinski, Benjamin 16 Jul '13

16 Jul '13

Dear freesurfer experts and users, we visualize white matter tracts of anorexia nervosa patients which are characterized by more movements in mri studies. We have the impression that we get worser TRACULA outputs by using bbregister than flt. Could it be possible that bbregister needs a higher resolution than flt? So summarizing it is more sensitive for motion? kind regards Benjamin

3 3

SWI acquisition question
by Fotiadis, Panagiotis 16 Jul '13

16 Jul '13

Hi FreeSurfer community, I am analyzing some SWI scans of some old subjects. For most of my subjects when a SWI was acquired, the resulting files that I would get would be a Magnitude image, a Phase image, a mIP image, and the SWI image. However, I have a couple of subjects for whom I get only the mIP and the SWI images, and the SWI image looks more blurry than it's supposed to be. I was wondering: 1) Why do I get only the mIP and the SWI images for those subjects and not the magnitude and phase images as well, 2) Is it normal for the mIP to look like that? Usually it looks different, and 3) Why is the final SWI image more blurry than usual? By looking at the image, I believe that maybe the head coil was not properly connected to the scanner at the time of the acquisition but I don't know whether that would be a sufficient explanation. I have attached two attachments: One of the resulting mIP and one of the equivalent blurry SWI of the same subject. Thank you in advance for your help and time! Best, Panos

1 0

Motion correction in recon-all
by Fotiadis, Panagiotis 16 Jul '13

16 Jul '13

Hi FS Community, I was wondering which command(s) is(are) being used in the "Motion Correction" part of the recon-all process, to merge the mri/orig/XXX.mgz files and correct for any motion. Thank you for your time! Best, Panos

2 2

Fwd: subcortical segmentation
by Liane Hunter 16 Jul '13

16 Jul '13

---------- Forwarded message ---------- From: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> Date: Tue, Jul 16, 2013 at 12:44 PM Subject: Re: [Freesurfer] subcortical segmentation To: Liane Hunter <liane.hunter(a)med.einstein.yu.edu> Hi Liane can you cc the list so others can answer? You will get better and faster answers that way, and also others will be able to learn from the replies. thanks Bruce On Tue, 16 Jul 2013, Liane Hunter wrote: Thanks so much.As as clarification: should I make edits to the > brainmask.mgz > > or norm.mgz file. > Likewise, is this the correct command to run to fix edits to the > subcortical > segmentation: > > normalization -maskbfs -segmentation -autorecon2-wm -autorecon3. > > Thank you again for all of your help > Liane > > > > On Tue, Jul 16, 2013 at 11:40 AM, Bruce Fischl <fischl(a)nmr.mgh.harvard.edu > > > wrote: > Hi Liane > the brainmask is not for subcortical segmentation - use the > norm.mgz if you want to visualize subcortical structures > > cheers > Bruce > On Tue, 16 Jul 2013, Liane Hunter wrote: > > > Hello, > I noticed that when I am checking the subcortical > segmentation the brainmask does not accurately display the > subcortical structures > ( ie. the putamen is too small). The segmentation appears > to be correct when I compare it to the original > file. Does anyone know why this is the case? > Also, if I make changes to the subcortical segmentation > based on the original file will this change accurately > edit the brainmask? > > Thank you. > Sincerely, > Liane > > Liane HunterMD/Ph.D Candidate > Albert Einstein College of Medicine > > > > > 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<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. > > > > > -- > Liane HunterMD/Ph.D Candidate > Albert Einstein College of Medicine > > -- Liane Hunter MD/Ph.D Candidate Albert Einstein College of Medicine

1 0

Button Press Regressor
by Nicholas Paul Denisuk 16 Jul '13

16 Jul '13

Dear Freesurfer Users, I have opened up the X.mat file from the analysis. I think that the button regressor is the third column in the file, but I am not 100% sure. Do you know which column the button regressor is? Below is the background info to this question. >>I am trying to regress out the activity associated with button presses in my GLM analysis. How do I generate a nuisance regressor file analogous to a white matter regressor generated in functional connectivity (a one column time course file)? Thanks you, Nick >>Hi Nick, please remember to post to the list and not to us personally. To answer your question, you could create an run an analysis with just the button press. Go into the analysis and get the button press regressor from X.mat, then save it as a text file in the run folder. Then make your analysis and specify this as a task regressor. I *think* that should work. doug

2 1

Segfault error
by Rune Frandsen 16 Jul '13

16 Jul '13

I have just installed Freesurfer and wanted to try tkmedit I get a Segfault error and it asked me to write to this mail. The .xdebug_tkmedit is tkmedit started: Wed Jul 10 10:53:23 2013 /usr/local/freesurfer/tktools/tkmedit.bin $Id: tkmedit.c,v 1.343 2011/03/01 01:41:22 nicks Exp $ $Name: stable5 $ Segfault Creating main window xDebug stack (length: 2) 01: main() 01: Creating main window 00: 00: My system is a fresh install of CentOS 6.4 64bit installed in VirtualBox. The only other program installed is FSL. The base machine is a HP Envy dv6 i7 16GB with Win8 64bit. Is this a known problem? Best regards Rune Frandsen (MD, phd student) Danish Center for Sleep Medicine runefrandsen(a)gmail.com

3 4

mritotal vs. talairach_avi scripts
by Mark Plantz 16 Jul '13

16 Jul '13

Hello, I have a quick question about talairach transform. I know that FreeSurfer previously used mritotal to create the linear talairach transform. You could check the "final objective function value" in the mritotal.log file. I think the ideal value was in the ballpark of 0.06. Is there a way to check for this value in the current talairach_avi script? If so, where would it be found? Possibly in the recon-all.log? Thank you for the help. - Mark

1 0

← Newer
1
...
13
14
15
16
17
18
19
...
34
Older →

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 July 2013