Freesurfer

freesurfer@nmr.mgh.harvard.edu

14 participants
30096 discussions

Re: [Freesurfer] convert labels to volumetric masks
by limengsecret 25 Apr '13

25 Apr '13

Hi Doug, Thanks for your reply. The mask created is used to register with the normalized functional data and to define the roi, so could I run the following command? tkregister2 -mov sub/func.nii -noedit -s fsaverage -regheader -reg ./register.dat/ and the func.nii has been normalized (sub.nii) to the MNI space. The .w file I download from http://ltl.tkk.fi/wiki/Atlas, indicates the maximum probability map. Best wishes, Meng Dear FreeSurfers, I want to convert cortical labels to volumetric masks, and I run the following commands: 1. /tkregister2 -mov sub/mri/rawavg.mgz -noedit -s sub -regheader -reg ./register.dat/ 2. /mri_label2vol -label ./lh.entorhinal.label -temp sub/mri/rawavg.mgz -subject sub -hemi lh -o sub_lh_entorhinal.nii.gz -proj frac 0 1 .1 -fillthresh .3 -reg ../mri/register.dat/ if I do not run the recon-all command, could I change the rawavg.mgz to sub.nii (original data)? What is sub.nii? The second problem is that when I convert the annotation file to volumetric mask, whether I should convert the file to labels using mri_annotation2label first, then run the commands as decribed above, or there is a way that could convert the annotation file to volumetric mask directly? You can use the same label2vol command above but spec --annot instead of --label The last problem is that how to convert the .curv file and .w file to volumetic mask? mri_surf2vol should work for the curv file. Why do you have a .w file? BTW, for our standard annotations (aparc and aparc.a2009s) you will probably be happier with the results if you use mri_label2vol with --seg aparc+aseg.mgz or --seg aparc.a2009s+aseg.mgz. Otherwise, there are some holes that are created if you use the surface labels. doug

1 0

Re: [Freesurfer] mri_glmfit --no-sim error: cannot find any csd files
by Jeni Chen 25 Apr '13

25 Apr '13

The original command line: > mri_glmfit-sim \ > --glmdir GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS \ > --cwpvalthresh .999 \ > --no-sim GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS/csd/mc-z.abs.3.j001-1GroupAvg-Cor-thickness-Mod-VAS.csd To which you suggested > mri_glmfit-sim \ > --glmdir GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS \ > --cwpvalthresh .999 \ > --no-sim 1GroupAvg-Cor-thickness-Mod-VAS And it did not work either. Thanks. jeni Douglas N Greve Thu, 25 Apr 2013 14:08:32 -0700 > what was your orignal mri_glmfit-sim command line? On 04/25/2013 05:17 PM, Jeni Chen wrote: > Hi Doug, > > Sorry; didn't realize it wasn't included. > > So to state the original problem, I tried running mri_glmfit-sim with > the following options (previous message included below): > > mri_glmfit-sim \ > --glmdir GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS \ > --cwpvalthresh .999 \ > --no-sim > GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS/csd/mc-z.abs.3.j001-1GroupAvg-Cor-thickness-Mod-VAS.csd > > I get an error message saying "cannot find any csd files". You > suggested putting just the csd base > (i.e. 1GroupAvg-Cor-thickness-Mod-VAS) for the --no-sim flag, which I > did, but still gave the same error message. I've attached the error > log if you want to have a look. > > Thanks. > > Jeni > > > > >> Hi Jeni, what is your command line and terminal output? In general, >> just saying that something does not work is not enough information to >> track down the problem. Also, please include previous correspondence. >> We answer a lot of emails here, and the easier you make it on us, the >> faster we can solve your problem. thanks! doug > On 04/25/2013 04:41 PM, Jeni Chen wrote: > > Thansk Doug. I tried it but still doesn't work. Any other thoughts? > > > > jeni > > > > > > _______________________________________________ > > Freesurfer mailing list > >Freesurfer(a)nmr.mgh.harvard.edu <mailto:Freesurfer@nmr.mgh.harvard.edu> > >https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > >> Hi Jeni, done pass that path to the csd, just the csd base. In your >> case I think it would be 1GroupAvg-Cor-thickness-Mod-VAS >> doug >>> On 4/24/13 8:47 PM, Jeni Chen wrote: >> >> Hello, >> >> I was trying to run mri_glmfit-sim at a lower cwpvalthresh so I >> specified the --no-sim option to avoid re-running the entire >> simulation again. I entered my DSC file but in the end I get the >> error message: "cannot find any csd files". Here is what I entered: >> >> mri_glmfit-sim \ >> --glmdir GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS \ >> --cwpvalthresh .999 \ >> >> --no-sim >> >> GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS/csd/mc-z.abs.3.j001-1GroupAvg-Cor-thickness-Mod-VAS.csdI >> copied and pasted the csd file from its own folder to the top >> directory, but it gave the same error. I've attached the error >> log in case you want to have a look. >> >> Thanks for your help! >> >> Jeni >> > > > > _______________________________________________ > Freesurfer mailing list > Freesurfer(a)nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

2 5

-autorecon2-pial
by obregon＠nmr.mgh.harvard.edu 25 Apr '13

25 Apr '13

Hi, I was wondering if the -autorecon2-pial flag has been superseded on the newer release of Freesurfer, for it isn't being recognized? Also, is the use of this flag equivalent to running recon-all with the -autorecon-pial option? Thanks, Gabriel

2 1

Re: [Freesurfer] mri_glmfit --no-sim error: cannot find any csd files
by Jeni Chen 25 Apr '13

25 Apr '13

Hi Doug, Sorry; didn't realize it wasn't included. So to state the original problem, I tried running mri_glmfit-sim with the following options (previous message included below): mri_glmfit-sim \ --glmdir GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS \ --cwpvalthresh .999 \ --no-sim GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS/csd/mc-z.abs.3.j001-1GroupAvg-Cor-thickness-Mod-VAS.csd I get an error message saying "cannot find any csd files". You suggested putting just the csd base (i.e. 1GroupAvg-Cor-thickness-Mod-VAS) for the --no-sim flag, which I did, but still gave the same error message. I've attached the error log if you want to have a look. Thanks. Jeni > Hi Jeni, what is your command line and terminal output? In general, just > saying that something does not work is not enough information to track > down the problem. Also, please include previous correspondence. We > answer a lot of emails here, and the easier you make it on us, the > faster we can solve your problem. > thanks! > doug On 04/25/2013 04:41 PM, Jeni Chen wrote: > Thansk Doug. I tried it but still doesn't work. Any other thoughts? > > jeni > > > _______________________________________________ > Freesurfer mailing list > Freesurfer(a)nmr.mgh.harvard.edu > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > > Hi Jeni, done pass that path to the csd, just the csd base. In your case I think it would be 1GroupAvg-Cor-thickness-Mod-VAS > doug > > >> On 4/24/13 8:47 PM, Jeni Chen wrote: > Hello, > > I was trying to run mri_glmfit-sim at a lower cwpvalthresh so I specified the --no-sim option to avoid re-running the entire simulation again. I entered my DSC file but in the end I get the error message: "cannot find any csd files". Here is what I entered: > mri_glmfit-sim \ > --glmdir GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS \ > --cwpvalthresh .999 \ > --no-sim GLMFIT/LH_1GroupAvg_ModVAS_nuissanceAGE_DOSS/csd/mc-z.abs.3.j001-1GroupAvg-Cor-thickness-Mod-VAS.csd > I copied and pasted the csd file from its own folder to the top directory, but it gave the same error. I've attached the error log in case you want to have a look. > Thanks for your help! > > Jeni

2 1

Re: [Freesurfer] mri_glmfit --no-sim error: cannot find any csd files
by Jeni Chen 25 Apr '13

25 Apr '13

Thansk Doug. I tried it but still doesn't work. Any other thoughts? jeni

2 1

HeadSurface to MEG fiducials
by José Ángel Pineda 25 Apr '13

25 Apr '13

Dear expert, I am doing source reconstruction with MEG. First, I am corregistering my head surface mesh from the freesurfer (obtained with mkheadsurf) to the headshape of the subject obtained with the Polhemus tracking system. So I already have one affine transformation (Aff_1) The head surface was obtained over the T1 image in voxel native space (not orig.nii) The next step is to apply an affine transformation to the T1 to bring it into the same coordinates as the headshape (and fiducials). Then I apply Aff_1 to the T1 and check the results. I observe there is an offset displacement between the T1 and the fiducial points. I dont know what to do to solve this. I read that as my head surface is in RAS space I should consider that, but Ive tried to correct this introducing the affine tranformation matrices that are mentioned on the web for Coordinates in Freesurfer, but Im not able to correct it. Could you please give me any advice? Thanks Jose -- José Angel Pineda Pardo Center of Biomedical Technology (CTB) Parque Científico y Tecnológico de la UPM Campus de Montegancedo 28223 Pozuelo de Alarcón, Madrid, ES

2 1

mri_ms_fitparms_questions
by dgw 25 Apr '13

25 Apr '13

I have run mri_ms_fitparms (version 5.1.0) on phantom data from two multiecho flash volumes (5 and 20 degrees with 8 echos: 1.67ms - 14.83ms in 1.88ms steps) at 7T utilizing a b1 map for correction. How does mri_ms_fitparms know the field strength (I can't figure out where the information is when I read the mgz files using MRIread.m)? I notice that the T1 volume seems to have graded steps of 5ms. Is this due to the table of 1ms stepped T1 values in the table described in Fischl et al. 2004 (and the data in this case being collected at 7T)? Is there any way to increase the sensitivity? Interestingly, this effect seems to only have a small effect on the values of the PD.mgz; however, from the 2004 Fischl paper, I would have expected it to have a large effect. What concerns should I have about using the PD.mgz to represent the Equilibrium Magnetization in terms of accuracy of the values? Thank You, Dan

2 2

Correcting bad segmentation/WM surfacing
by Jonathan Holt 25 Apr '13

25 Apr '13

See attached image for reference. Large area has been incorrectly segmented as WM, also WM surface has completely been skipped over. Planning to remove segmentation in that area and run recon-all again, is this correct or should something else be done to alleviate issue. Worth noting this issue is only in one hemisphere.

2 1

convert labels to volumetric masks
by limengsecret 25 Apr '13

25 Apr '13

Dear FreeSurfers, I want to convert cortical labels to volumetric masks, and I run the following commands: 1. tkregister2 �Cmov sub/mri/rawavg.mgz �Cnoedit �Cs sub �Cregheader �Creg ./register.dat 2. mri_label2vol �Clabel ./lh.entorhinal.label �Ctemp sub/mri/rawavg.mgz �Csubject sub �Chemi lh �Co sub_lh_entorhinal.nii.gz �Cproj frac 0 1 .1 �Cfillthresh .3 �Creg ../mri/register.dat if I do not run the recon-all command, could I change the rawavg.mgz to sub.nii (original data)? The second problem is that when I convert the annotation file to volumetric mask, whether I should convert the file to labels using mri_annotation2label first, then run the commands as decribed above, or there is a way that could convert the annotation file to volumetric mask directly? The last problem is that how to convert the .curv file and .w file to volumetic mask? Any help will be very appreciated. Best wishes, Meng

2 1

System recommendations for Freesurfer
by Graham Wideman 25 Apr '13

25 Apr '13

Hi all, I am looking for current recommendations for specs for a desktop Linux box for running Freesurfer (and friends), both recon-all, and the interactive graphical tools. I have read the generic recommendations here: http://surfer.nmr.mgh.harvard.edu/fswiki/SystemRequirements (also here as surfer is down)... http://webcache.googleusercontent.com/search?q=cache:Q113A-rSkxEJ:surfer.nm… Feel free to point me to existing email threads if I've missed the place where this was definitively covered! I'd like any more detailed input such as: CPU: Intel vs AMD, and the extent to which more cores increase performance. Video card: -- AMD/ATI or nVidia? My model is that nVidia has had better support for OpenGL/Mesa -- still true? -- if nVidia, then how many CUDA cores and RAM is beneficial for FS's CUDA code? Hard drive: Does SSD make much difference while running recon-all? Linux distro: Which are favored for use with Freesurfer, FSL, etc? Thanks! -- Graham

3 3

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