Freesurfer

freesurfer@nmr.mgh.harvard.edu

17 participants
30086 discussions

Account password on Virtualbox VDI
by Mcnorgan, Christopher 18 Sep '20

18 Sep '20

External Email - Use Caution Hi, I have a couple students who are windows users who I am training remotely. They’ll be using the VDI found at: https://surfer.nmr.mgh.harvard.edu/fswiki/VM_67 I’ve never tried this and can give no guidance, as I have no Windows computers at my disposal. A student tried this and said it had a default ‘developer’ user, and asked what the password was for this account. A google search revealed this question has been posted in the past, with a reply that the password would be sent via email. /********************************************** * 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 **********************************************/

1 0

voxel to ras transform written by mri_convert -c
by Billah, Tashrif 18 Sep '20

18 Sep '20

Hi there, How is the rightmost column in `mri_convert -c` calculated? I have the following fslhd (relevant fields shown only) in a NIFTI image: ``` dim1 208 dim2 300 dim3 320 qto_xyz:1 0.799690 0.022176 0.001352 -86.687996 qto_xyz:2 -0.021442 0.783122 -0.162051 -62.728546 qto_xyz:3 -0.005816 0.161952 0.783414 -174.620270 qto_xyz:4 0.000000 0.000000 0.000000 1.000000 ``` After `mri_convert -c`, mri_convert reports the following (relevant fields shown only) in the MGZ image: ``` voxel to ras transform: -1.0000 -0.0000 0.0000 128.0225 -0.0000 0.0000 1.0000 -101.4184 0.0000 -1.0000 -0.0000 102.4140 0.0000 0.0000 0.0000 1.0000 ``` Shouldn't the right-most column be: 128.5 -128.5 128.5 1 The numbers are (256+1)/2=128.5 Best, Tashrif

2 1

Autorecon after control points error
by Eleni Nikalexi 18 Sep '20

18 Sep '20

External Email - Use Caution Dear Freesurfer developers, After we run recon-all in a subject we noticed some mistakes, so we added manually some control points in a point set we named control.dat. Then we tried to run autorecon, with the command: recon-all -autorecon2-cp -autorecon3 -subjid Sub42 but we stumbled upon the following error message: using control points from file /project/data/scans/subject/Sub42/tmp/control.dat... assuming input volume is MGH (Van der Kouwe) MP-RAGE using segmentation for initial intensity normalization using MR volume brainmask.mgz to mask input volume... reading from norm.mgz... Reading aseg aseg.presurf.mgz normalizing image... Reading 133 control points... only using 133 control points from file, mean 69.1, scaling by 1.59... building Voronoi diagram... performing soap bubble smoothing, sigma = 8... processing with aseg removing outliers in the aseg WM... 580 control points removed Building bias image building Voronoi diagram... performing soap bubble smoothing, sigma = 0... Smoothing with sigma 8 Applying bias correction HISTOalloc(-2147483648): could not allocate histogram building Voronoi diagram... performing soap bubble smoothing, sigma = 8... Iterating 2 times --------------------------------- 3d normalization pass 1 of 2 white matter peak found at 110 Cannot allocate memory Linux hpc137 2.6.32-696.16.1.el6.x86_64 #1 SMP Wed Nov 15 16:51:15 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux recon-all -s Sub42 exited with ERRORS What do you think is the problem? Thank you in advance! Best regards, Eleni

2 1

Resting state fMRI preproc-sess registration error
by Wenzhen Zhao 18 Sep '20

18 Sep '20

External Email - Use Caution Dear freesurfer developers, I am working on preprocessing resting state fMRI and used the following command. $ preproc-sess -s sess01 -fsd rest -stc siemens -surface fsaverage lhrh -surf-fwhm 5 -mni305-2mm -vol-fwhm 0 -per-session When I tried to visualize f.nii.gz with registration file applied and overlaid with aparc+aseg.mgz from structural analysis, they do not overlap. It seems like the orientation is not the same, and the f.nii.gz with registration seems to be 90 degree rotated from apjarc+ aseg. mgz. (I wish I could attach a screenshot, but last time I tried, it says my email exceeded the capacity and sent to the moderator) I also tried bbregister. I was suggested not to use —init-rr, so I tried all the other registration methods, but same problem occurs. $bbregister --mov sess01/rest/001/template.nii.gz --t2 --s subject --lta register.dof6.lta —init-rr does a good job when I checked visually by applying registration on f.nii.gz and checked if it overlaps with structural scan. But, I thought the registration file goes both ways, but when I checked structural scan (orig.mgz or aparc+aseg.mgz) with registration file and overlapped with functional scan (f.nii.gz), it does not overlap. Is there a specific way that I have to apply a registration file to visualize the transformation in the opposite direction? If preprocess registration does not work, is there any other method that I can try to make them aligned? Thank you for your time and help ! Best, Wenzhen

3 4

Fwd: non-white matter hypointensity
by L 18 Sep '20

18 Sep '20

External Email - Use Caution Dear experts, I am wondering what the non-white matter hypointensity refers to. Does it refer to hypointensity in subcortical regions or all brain regions except white matter? When we do data analysis using volume results from ROI tables, we found a greater volume of non-wm hypointensity in depressed subjects with- than without high vascular risk factors. However, I am not sure where the non-white matter hypointensity ROI covers, and it is possible to visually display the non-wm hypointensity regions in freeview? Appreciate your guidance in advance. Lihong

3 3

Converting Volume ROI to Surface .annot file
by Paul Dhami 18 Sep '20

18 Sep '20

External Email - Use Caution Dear Freesurfer community, apologizes in advance for what may be a simple question. I have a volume region of interest file in .nii format. I would like to do a surface functional connectivity analysis in a MATLAB toolbox, and to do so, I would need to somehow convert this volume ROI to an .annot file. I was wondering if Freesurfer has any tools to do this? Could it be accomplished using a combination of mri_vol2label and then mri_label2annot? Best, Paul

2 1

mri_nu_correct.mni, mri_normalize extremely dark output
by Caspar M. Schwiedrzik 18 Sep '20

18 Sep '20

External Email - Use Caution Hi! I am having some issue at the interface between mri_nu_correct.mni and mri_normalize, using v5.3. Specifically, the output when applying mri_nu_correct to a T1 is quite dark, with white matter in the range of 15. Contrast is ok. When I apply mri_normalize to nu, it does not scale the data to white matter 110. This happens with and without a brain mask, and with various other parameter choices. I am tempted to just add 100 to all voxels. Any advice? Thank you, Caspar

3 3

mri_surf2surf template rh.white to fsaverage5
by 1248742467 18 Sep '20

18 Sep '20

External Email - Use Caution hello,Fs Because I want to build an N*N matrix based on the fixed-point number N of rh.white, but the original rh.white has too many fixed-point numbers, a memory error occurs. I saw a paper, which can be registered to the fsaverage5 space, and then the number of vertices can be reduced to 10242. I learned that the mri_surf2surf command is fine, but how should I use this command? If you can please give me an example Looking forward to your reply

2 1

Recon-all crooked volumes
by Hasimoglu, Yasemin 18 Sep '20

18 Sep '20

Hello, I recently ran recon-all on a group of subjects with FreeSurfer 7.1.0, and for one subject some of the resulting volumes look crooked. The 001.mgz volume looks normal, but every other volume such as orig.mgz, T1.mgz, brainmask.mgz have a severe tilt to them. When we open surfaces on top of volumes, their orientations follow whichever volume is loaded first. I've attached 2 sample screenshots here. We suspected that this could be related to a Talairach registration issue and inspected the volumes using the following command: tkregister2 --mgz --s $sub --fstal --surf orig However, everything seems to be aligned correctly on here, so we're not sure if there's anything we can do. Is this something that needs additional correction, or can we use this data in analysis as is? I've also attached the recon-all.log in case it's helpful. Best, Yasemin

2 1

Prescan normalize and freesurfer
by Mikolas, Pavol 18 Sep '20

18 Sep '20

External Email - Use Caution Dear Freesurfer Developers, I am analyzing the T1 Data of a multicentric study (N=309, 6 Centers) using Freesurfer. I found out, that most of the scans were performed with the 'prescan normalize' option. However, by approx.. 12 % of participants, this option is missing. I also have complete data of all participants WITHOUT prescan normalize. In this analysis, we are interested in the correlation of GM volume in specific ROIs with clinical symptoms. Now we have two options: 1. Use the 'prescan normalize' dataset and account for the 12% without this option as a covariate 2. Use the complete dataset without 'prescan normalize' It would be very helpful, if you could share your arguments / experience with this, so that we proceed in the best manner. Thank you, best regards Pavol

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