Freesurfer

freesurfer@nmr.mgh.harvard.edu

30132 discussions

Volume-defined roi
by Erik Lindberg 27 Mar '20

27 Mar '20

External Email - Use Caution Hi I am reading your tutorial about how to extract Cortical Thickness of a volume-defined ROI Wonder if there is a script that would run these steps in a bach particularly the steps that needs to be done on each individual subjects: map subject thickness data to the fsaverage and run mri_segstats to extract mean thickness Best Olof

2 1

Fwd: recon-all-lh error : cannot allocate memory
by Jong-eun Lee 27 Mar '20

27 Mar '20

External Email - Use Caution Hello FreeSurfer Developers, I'm trying to test Fmriprep pipeline for HBN datasets. But I got an error: CORRECTING DEFECT 158 (vertices=200262, convex hull=34956, v0=308501) Excessive topologic defect encountered: could not allocate -993586868 edges for retessellation Cannot allocate memory Linux cnir14 2.6.32-754.23.1.el6.x86_64 #1 SMP Thu Sep 26 12:05:41 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux recon-all -s sub-NDARAC853DTE exited with ERRORS at Mon Mar 23 18:37:34 KST 202 I attach recon-all-lh log file. ᐧ ᐧ

2 1

selxavg3-sess not working
by Nasiriavanaki, Zahra 27 Mar '20

27 Mar '20

Hi Freesurferers I am trying to run selxavg3-sess for a single subject (3T fMRI data). Oddly, the terminal output shows that it doesn't need updating for preprocessing and ends with this: "Started at Thu Mar 12 10:48:09 EDT 2020 Ended at Thu Mar 12 10:48:12 EDT 2020 preproc-sess done". It seems that it just checks the preprocessing, but does not initiate the selxavg analysis. There is no error neither selxavg log file. my command: selxavg3-sess -s $subj -d ../all_subjects -analysis loom.lh -force I appreciate any comments on this issue. Thanks Mona Zahra (Mona) Nasiriavanaki Postdoctoral Research Fellow Martinos Center for Biomedical Imaging Massachusetts General Hospital 149 13th Street, 149-2615 Charlestown, MA, USA, 02129

2 5

Permutation with Non-Ortohgonal Matrices
by Hua, Jessica 27 Mar '20

27 Mar '20

External Email - Use Caution Hi FreeSurfer Experts, I am trying to correct for multiple comparisons by running a permutation for a non-orthogonal design. Based on reading through the blog, I came up with the follow script: mri_glmfit-sim --glmdir rh.volume.AgeSex.10.glmdir --sim perm-resid 1000 3.0 --sim-sign abs --cwp 0.05 --2spaces I am getting the following error message: ERROR: Flag abs unrecognized. And when I tried to put the "abs" after "1000 3.0," I got the error message saying that I needed to specify the --sim-sign. Could you please let me know how I need to edit my script? I have already run mris_preproc and mri_glmfit. Thanks! --- Jessica Hua, M.A. Cognitive and Emotional Control Lab Doctoral Candidate Clinical Psychology University of Missouri

2 1

Registration file types confusion
by Xiaojiang Yang 27 Mar '20

27 Mar '20

External Email - Use Caution Hi everyone, I am confused by the two types of registration files: .lta and .dat. .dat file is simpler and easier to understand. But for the .lta files, there seems to be no detailed explanation regarding it. Can anyone point out to me where can I find the detail explanation of the file? And, what's the relationship between .lta and .dat file? Also, .lta files seem to cause some compatibility issues. For example, if I use the tkregister2 utility to make some changes to the registration file of type .lta (generated by bbregister), like below tkregister2 --mov $someSeries --reg $SUBJECTS_DIR/$subjectid/mri/transforms/FLAIR.auto.lta --surf pial and save it from the GUI, then many errors occur in the console window, like below: ...... MatrixAsciiReadFrom: could not scanf parms ERROR: TransformLTA2RegDat(): type = 2011264000, must be 1 or 0 or 14 Segmentation fault (core dumped) Actually that is because the registration file has been saved in the format of .dat, although the file extension is still .lta, and the tkregister2 program finally cracks when reading the newly saved .lta file. Anyone can solve my confusion? Thank you!

2 1

Using mri_glmfit to regress out wm and ventricle signal
by Mcnorgan, Christopher 27 Mar '20

27 Mar '20

External Email - Use Caution I’ve been using mri_glmfit to regress out motion correction parameters and linear trends and saving residuals in surface space by first generating a QA matrix for the volume: mri_glmfit --y ${source_dir}/${r}/${filepat}.${surf}.${hemi}.nii.gz \ --glmdir ${source_dir}/${r}/{hemi}.detrend \ --qa \ --surf ${SURFTOUSE} lh Then merging the QA and MC matrices into a single matrix of nuisance regressors: paste ${source_dir}/${r}/lh.detrend/Xg.dat mcprextreg > nuisance And finally, regressing out the nuisance values and saving the residuals mri_glmfit --y ${source_dir}/${r}/${filepat}.${surf}.${hemi}.nii.gz \ --glmdir ${source_dir}/${r}/${hemi}.detrend \ --X nuisance \ --no-contrasts-ok --save-yhat --eres-save \ --surf ${SURFTOUSE} ${hemi} Is there a straightforward way of obtaining mean white matter and csf signal in my nuisance regressor matrix? I was unable to find any preprocessing step or application of mri_glmfit that will provide these values or regress out these values automatically. (Also note: the —qa flag is not documented in the mri_glmfit help; I must’ve learned of it from a previous email to this list) Thanks /********************************************** * Chris McNorgan * Assistant Professor * Department of Psychology * University at Buffalo, * The State University of New York * http://ccnlab.buffalo.edu/ * Office: 716.645.0236 * Lab: 716.645.0222 **********************************************/

2 1

Re: [Freesurfer] PetSurfer reference region
by Dincer, Aylin 27 Mar '20

27 Mar '20

External Email - Use Caution Hello PetSurfer experts, We found the SUVR values for the reference region (lh/rh cerebellum cortex) does not equal 1 in the gtm.stat.dat file. We tried a few things: 1) leaving the PET image as multi-frames (ref region SUVR about 1.34) 2) concatenating the frames as suggested in the previous response (ref region SUVR about 1.34) 3) running with the SUV image (ref region SUVR about 1.003). I attached the terminal output and the gtm.stat.dat file from the concatenated image run. Here is the command we ran using v6: mri_gtmpvc --i regmean.nii.gz --reg regmean_template.reg.lta --psf 8--seg ${MR_ID}/mri/gtmseg.mgz --default-seg-merge --auto-mask PSF+2 .01 --mgx .01 --rescale 8 47 -o regmean_gtmpvc_output Let me know if you have suggestions. Thanks so much, Aylin _______________________ Message: 11 Date: Fri, 6 Mar 2020 16:59:53 -0500 From: "Douglas N. Greve" <dgreve(a)mgh.harvard.edu> Subject: Re: [Freesurfer] PetSurfer reference region To: <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <8c262861-3272-151b-ae18-9068984084d4(a)mgh.harvard.edu> Content-Type: text/plain; charset="windows-1252" I think the issue is that your input has multiple frames (6). It will compute the scaling from all 6 whereas the gtm.stats file is only from the first frame. Try running it with a single frame. Note that the gtm.nii.gz will have all the frames for each ROI. On 3/6/2020 3:56 PM, Dincer, Aylin wrote: > External Email - Use Caution > > Thanks for the quick reply. I attached the terminal output to this email. We are running this on v6. > > Best, > Aylin > > _______________________ > Message: 9 > Date: Fri, 6 Mar 2020 11:50:51 -0500 > From: "Douglas N. Greve" <dgreve(a)mgh.harvard.edu> > Subject: Re: [Freesurfer] PetSurfer reference region > To: <freesurfer(a)nmr.mgh.harvard.edu> > Message-ID: <64d6089e-f818-e30e-deff-98a812b72636(a)mgh.harvard.edu> > Content-Type: text/plain; charset="windows-1252" > > I cannot replicate this in either v6 or v7beta. Can you send the > terminal output? > > On 3/5/2020 1:32 PM, Dincer, Aylin wrote: >> ????????External Email - Use Caution >> >> Hello PetSurfer experts, >> >> We are using the cerebellum cortex as our reference region, but >> noticed that the SUVR values for lh/rh cerebellum cortex do not have a >> value of 1 (which is what we were expecting). The actual SUVR values >> are 1.31 for the lh and 1.35 for the rh cerebellum cortex. I attached >> the stats file (gtm.stats.dat) and below is the command we ran. >> >> mri_gtmpvc --i regmean.nii.gz --reg regmean_template.reg.lta --psf >> ${FWHM} --seg ${MR_ID}/mri/gtmseg.mgz --default-seg-merge --auto-mask >> PSF+2 .01 --mgx .01 --rescale 8 47 ?o regmean_gtmpvc_output >> >> Let me know if you need any more details. >> >> Thanks so much, >> >> Aylin >> >> ???????????????????? >> >> _____________________________ >> >> Aylin Dincer >> >> Senior Research Technician >> >> NeuroImaging Laboratories >> >> Imaging Core, Benzinger Lab >> >> 314-362-3629 ________________________________ The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying 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 mail.

2 1

Density of the skull
by Pazuelo, Jeremy Raphael Judah 27 Mar '20

27 Mar '20

Hello FreeSurfer Developers, For my research on neurostimulation using ultrasound, I am trying to obtain a 3D matrix with the density values of a patient's skull. Is there a way with FreeSurfer to obtain a file containing the density of the skull. I could then transform this file on Matlab to obtain the matrix. Thank you for your help, Jeremy Pazuelo

2 1

Run FreeSurfer recon-all on isoDWI
by Yixin Ma 27 Mar '20

27 Mar '20

External Email - Use Caution I was trying to use the average of DWI images (isoDWI) to run recon-all, and it didn't work. I understand that FreeSurfer can only take a T1-weighted image for now. But how hard would it be to modify the code by myself to make it work? If it works, how accurate do you think it would be to run FreeSurfer on isoDWI? All the best and thank you for your help, Yixin Message: 8 Date: Fri, 13 Mar 2020 18:44:18 -0400 From: "Douglas N. Greve" <dgreve(a)mgh.harvard.edu> Subject: Re: [Freesurfer] FreeSurfer Cortical Parcellation using isoDWI image only? To: <freesurfer(a)nmr.mgh.harvard.edu> Message-ID: <14b89d9d-130f-08f0-d593-71f2cdef3107(a)mgh.harvard.edu> Content-Type: text/plain; charset="utf-8" you are trying to run a DWI image through recon-all? Sorry, it can only take T1w right now On 3/12/2020 3:28 PM, Yixin Ma wrote: > > ????????External Email - Use Caution > > Hi Freesurfer experts, > > I'm writing to ask if it's possible to run FreeSurfer Cortical > Parcellation by just using the isoDWI image. If it's possible, it will > reduce the scan time and avoid misregistration between T1 and DWI. > > I started from just input the masked isoDWI (mask generated from > MRtrix3) for -reconall command, and attached is a screenshot of the > image of NU.mgz, which shows a clear delineation of WM/GM matter. So I > think it's promising, although the recon-all stopped because of the > wrong contrast. > My ideas on tackling this problem are: > 1) generate a corresponding "fake" T1-weighted image by > either?adjusting the value directly or using a brain Atlas and > deformable registration to create one. > 2) Tweak some parameters in FreeSurfer. > > Any suggestions or comments on my ideas are appreciated. > > Thank you very much and all the best, > > Yixin Ma >

2 1

mri_deface download link not found
by D'Silva, Hailey 27 Mar '20

27 Mar '20

External Email - Use Caution Hello Freesurfer Experts! I'm trying to anonymize a number of T1s by using mri_deface, but am getting a "Not Found" error when I try to download the tools for it. Thanks for your help, Hailey

2 1

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