Freesurfer

freesurfer@nmr.mgh.harvard.edu

18 participants
30067 discussions

Are inflated surfaces vertices transformed in Freeview?
by Sarah Lavelle 12 Dec '22

12 Dec '22

External Email - Use Caution Hi Doug, Thank you for your help! Yes, I have followed the points using vertex indices but there were some that didn't match up. It could be because of the 1-based indexing. Where would I change this? I am using freesurfer_read_surface.m to read in the vertices and faces : function [vertices, faces] = freesurfer_read_surf(fname) % freesurfer_read_surf - FreeSurfer I/O function to read a surface file % % [vertices, faces] = freesurfer_read_surf(fname) % % Reads the vertex coordinates (mm) and face lists from a surface file. % Copyright (C) 2000 Darren L. Weber % History: 08/2000, Darren.Weber_at_radiology.ucsf.edu % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % There used to be a line here where ver was set to the version of % the file. Without it, "ver" reverts to a matlab defined structure % and generates an error. %fprintf('FREESURFER_READ_SURF [v %s]\n',ver(11:15)); if(nargin < 1) help freesurfer_read_surf; return; end %QUAD_FILE_MAGIC_NUMBER = (-1 & 0x00ffffff) ; %NEW_QUAD_FILE_MAGIC_NUMBER = (-3 & 0x00ffffff) ; TRIANGLE_FILE_MAGIC_NUMBER = 16777214; QUAD_FILE_MAGIC_NUMBER = 16777215; % open it as a big-endian file fid = fopen(fname, 'rb', 'b'); if (fid < 0), str = sprintf('could not open surface file %s.', fname); error(str); end fprintf('...reading surface file: %s\n', fname); tic; magic = freesurfer_fread3(fid); if (magic == QUAD_FILE_MAGIC_NUMBER), Nvertices = freesurfer_fread3(fid); Nfaces = freesurfer_fread3(fid); fprintf('...reading %d quad file vertices\n',Nvertices); vertices = fread(fid, Nvertices*3, 'int16') ./ 100 ; if (nargout > 1), fprintf('...reading %d quad file faces (please wait)\n',Nfaces); faces = zeros(Nfaces,4); for iface = 1:Nfaces, for n=1:4, faces(iface,n) = freesurfer_fread3(fid) ; end if(~rem(iface, 10000)), fprintf(' %7.0f',iface); end if(~rem(iface,100000)), fprintf('\n'); end end end elseif (magic == TRIANGLE_FILE_MAGIC_NUMBER), fprintf('...reading triangle file\n'); tline = fgets(fid); % read creation date text line tline = fgets(fid); % read info text line Nvertices = fread(fid, 1, 'int32'); % number of vertices Nfaces = fread(fid, 1, 'int32'); % number of faces % vertices are read in column format and reshaped below vertices = fread(fid, Nvertices*3, 'float32'); % faces are read in column format and reshaped faces = fread(fid, Nfaces*3, 'int32'); faces = reshape(faces, 3, Nfaces)'; else str = sprintf('unknown magic number in surface file %s.', fname); error(str); end vertices = reshape(vertices, 3, Nvertices)'; fclose(fid); fprintf('...adding 1 to face indices for matlab compatibility.\n'); faces = faces + 1; t=toc; fprintf('...done (%6.2f sec)\n\n',t); return ii) Also, I want to transform a section of the inflated surface into a *2D flatmap*, and import it into *MATLAB* while *maintaining the same VertexIDs* so I know which vertex on the flatmap corresponds to a vertex on a pial surface. Do you know how I can do this? I have created multiple flatmaps in Freeview by : 1. Creating a patch on lh.inflated in Freeview and saving it 2. Flattening the patch using mris_flatten() in terminal 3. Loading lh.inflated in Freeview and then loading the flattened patch which then creates a 2D flatmap However, I cannot figure out how to load this 2D flatmap into MATLAB. I've tried saving the result in (3) as a surface and then loading it into matlab with freesurfer_read_surf.m. This takes a long time and the result looks a bit distorted. It also does not conserve the VertexIDs. Do you have any recommendations? Many thanks again for your help, Sarah

2 1

Pial surface does not follow the gray boundary
by 白晨 12 Dec '22

12 Dec '22

External Email - Use Caution Hello FreeSurfer Developers, I'm attempting to analyze the T1 imaging of children aged 2–7 years, a problem with the segmentation occurred,the pial surface of many subjects did not follow the gray matter boundary, and large area of gray matter in right hemisphere was excluded from pial surface (as shown in pictures), we used control points to extend white matter in order to extend pial surface, but can't fully resolve the problem, Is there any other solution? I will be very grateful if you can give meadvice. Also, I've attached the files of quality checking(Internal surface and external surface) and recon-all.log of a subject in case it's of any use. My setup is the following: FreeSurfer version: freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c platform: Ubuntu 20.04.4 LTS uname -a: Linux shenkang-B560M-AORUS-ELITE 5.15.0-53-generic #59~20.04.1-Ubuntu SMP Thu Oct 20 15:10:22 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux (T1 voxel size is 0.5mm*1mm*1mm) Any help would be greatly appreciated! With kind regards, Chen Bai

2 1

Creating and analyzing combined ROIs, combining different FreeSurfer versions
by Trolle, Carl 12 Dec '22

12 Dec '22

Dear FreeSurfer team, 1. We're trying to focus our cortical analysis following suggestions here https://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg23939.html. The FS ROI tutorial page mentions that labels should be drawn or registered on the fsaverage template; could we then break the fsaverage aparc_annot atlas and combine the fsaverage labels of interest into our new label or should we break each individual subjects aparc_annot atlas and then create the combined ROI from the subject specific labels? I'm assuming that we either way will have to resample the combined label to each subject using mri_label2label? 2. When running the mri_glmfit-sim analysis later on, the thread above mentions not using the --cache option. Can we use "--mczsim 3 pos" instead for a one-sided test? 1. We've had some success in correcting incorrect parcellations using the mris_register and mri_ca_label functions. As we had some issues with mris_register in 7.1.1, we used the developer version for these steps, but 7.1.1 when re-running recon-all as this is the version we used for our study. Is it ok to combine versions like this? Thank you for your time! Kind regards, Carl Trolle Research fellow Spaulding Neuroimaging Lab Platform: linux-centos7_x86_64-7.1.1-20200723-8b40551 FreeSurfer version: 7.1.1

2 1

ERROR: fast_mat2vol
by stdp82＠virgilio.it 12 Dec '22

12 Dec '22

External Email - Use Caution I would extract the rs-fMRI timeseries from Yeo’s Network components. I’m running: preproc-sess -s ${1} -fsd rest -sliceorder up -surface self lhrh -fwhm 0 -per-runfcseed-config -wm -fcname wm.dat -fsd rest -pca -cfg wm.config fcseed-sess -s ${1} -cfg wm.configfcseed-config -vcsf -fcname vcsf.dat -fsd rest -pca -cfg vcsf.configfcseed-sess -s ${1} -cfg vcsf.config mkanalysis-sess -analysis fc.noseed.surf.lh -surface self lh -fwhm 0 -notask -nuisreg vcsf.dat 5 -nuisreg wm.dat 5 -polyfit 5 -nskip 4 -mcextreg -fsd rest -TR 1.67 -per-run -hpf 0.01 -lpf 0.1 -nowhiten mkanalysis-sess -analysis fc.noseed.surf.rh -surface self rh -fwhm 0 -notask -nuisreg vcsf.dat 5 -nuisreg wm.dat 5 -polyfit 5 -nskip 4 -mcextreg -fsd rest -TR 1.67 -per-run -hpf 0.01 -lpf 0.1 -nowhiten The line below produced this error: selxavg3-sess -s ${1} -a fc.noseed.surf.lh -no-con-ok -svres Saving rho1 ERROR: fast_mat2vol: szvol inconsistent with mat2d nv = 151008, size(mat,2) = 142954 run 2 t= 1.6reading data ... 0.518631Temporally filteringSaving rho1 ERROR: fast_mat2vol: szvol inconsistent with mat2d nv = 151008, size(mat,2) = 142954Found 0 voxels with corrected AR1 > 0.90Not WhiteningFound 8054 zero-valued voxelsComputing contrasts The res folder in fc.noseed.surf.lh or fc.noseed.surf.rh contains res-001.nii.gz 158KB res-002.nii.gz 158KB that cannot be opended. Stefano

2 1

pial surface error: ?h.pial surface excludes part of gray matter/brainmask
by Alessia Sepe 12 Dec '22

12 Dec '22

External Email - Use Caution Hi FreeSurfer Developers, I have a major problem with the reconstruction of the pial surface. The issue is the opposite to the one relate to the brainmask: the reconstruction excludes part of the gray matter. In attachment an example of the output after running -autorecon2-wm -noaseg, and the recon-all.log if needed. Does anyone have any thoughts on how to fix it? FreeSurfer version: v5.3.0 up to 'recon-all -s <subjid> -autorecon2-inflate1 -noaseg -cc-crs <x> <y> <z> -pons-crs <x> <y> <z>', v6.0.0 for manual editing on the wm and 'recon-all -s <subjid> -autorecon2-wm -noaseg -cc-crs <x> <y> <z> -pons-crs <x> <y> <z>'. Thank you. Best wishes, Alessia Sepe MSc | PhD Fellow KU Leuven Laboratory for Neuro- and Psychophysiology Department of Medicine ON II Herestraat 49, 3000 Leuven

2 1

Effect size
by Lyna Mariam El Haffaf 12 Dec '22

12 Dec '22

External Email - Use Caution Hi! I'm taking the liberty of writing another email on this subject, as I didn't get a reply to my previous email. My colleagues and I performed a general linear model with FreeSurfer mri_glmfit for whole brain analysis in intergroup comparison (g2v1) with a Monte-Carlo correction. I'd like to get the effect size of the analysis and I was wondering how could I get the eta squared or R2? I also saw that the gamma.mgh file was the contrast effect size. Could this be used as an effect size for this type of analysis? I believe an equation such as : pseudo_rsquared = (1 - model.llf / model.llnull) could be helpful but I wonder how to include this in the mri_glmfit script. Thank you in advance! Lyna El Haffaf University of Montreal

2 1

Questions for FreeSurfer 7.3.2 recon-all process INCLUDING SynthSeg
by Yoonho Hwang 11 Dec '22

11 Dec '22

External Email - Use Caution Dear Bruce Fischl, Hi, I'm Yoonho. I know that new features, such as SynthSeg, SynthStrip, segment_subfields (beta version), etc, based on deep learning have been added after the FreeSurfer 7.3.2. Here I have questions (I read papers related to SynthSeg, SynthStrip, segment_subfields). 1. *In FreeSurfer 7.3.2, the recon-all function includes the mri_synthse function. Does the recon-all process include SynthSeg by default? Likewise, does SynthStrip have the same process by default?* 2. There are already processed files with FreeSurfer 7.2.0, *can we apply new functions (e.g., SynthSeg, SynthStrip, etc) to them?* 3. From SynthSeg related papers (e.g., Billot et al, 2021 and 2022), in the following quoted text, "Therefore, there is a clear need for a fast, accurate, and reproducible automated method, for segmentation of brain scans of *any contrast and resolution*, and that can adapt to a wide range of subject populations", Synthseg is said to be able to process MRI data of any contrast and resolution. *Does this also apply to data taken from other manufacturers (Siemens, GE, Philips) and protocols (FSPGR, MPRAGE, etc)?* Also, there are many studies that attempt to harmonize MRI data obtained from various sites. *If there are different parts such as MRI data acquisition time, manufacturer, and protocol, the harmonization can be solved using the SynthSeg function of FreeSurfer 7.3.2. (In other words, can we simply proceed with the recon-all process?)* Thank you for reading my long question Please let me know if there is anything you don't understand. Looking forward to your feedback. Thank you very much. Sincerely, Yoonho -- *Hwang, Yoonho* *Research professor* *Institute of Human Genomic Study,* *College of Medicine, Korea University* *#123, Jeokgeum-ro, Danwon-gu, Ansan-si, Gyeonggi-do, **15355, Republic of Korea.* *Email: ken.lmstar11(a)gmail.com <ken.lmstar11(a)gmail.com>*

1 0

ERROR: fast_mat2vol: szvol inconsistent with mat2d nv
by stdp82＠virgilio.it 11 Dec '22

11 Dec '22

External Email - Use Caution I would extract the rs-fMRI timeseries from Yeo’s Network components. I’m running: preproc-sess -s ${1} -fsd rest -sliceorder up -surface self lhrh -fwhm 0 -per-run fcseed-config -wm -fcname wm.dat -fsd rest -pca -cfg wm.config fcseed-sess -s ${1} -cfg wm.config fcseed-config -vcsf -fcname vcsf.dat -fsd rest -pca -cfg vcsf.config fcseed-sess -s ${1} -cfg vcsf.config mkanalysis-sess -analysis fc.noseed.surf.lh -surface self lh -fwhm 0 -notask -nuisreg vcsf.dat 5 -nuisreg wm.dat 5 -polyfit 5 -nskip 4 -mcextreg -fsd rest -TR 1.67 -per-run -hpf 0.01 -lpf 0.1 -nowhiten mkanalysis-sess -analysis fc.noseed.surf.rh -surface self rh -fwhm 0 -notask -nuisreg vcsf.dat 5 -nuisreg wm.dat 5 -polyfit 5 -nskip 4 -mcextreg -fsd rest -TR 1.67 -per-run -hpf 0.01 -lpf 0.1 -nowhiten The line below produced this error: selxavg3-sess -s ${1} -a fc.noseed.surf.lh -no-con-ok -svres Saving rho1 ERROR: fast_mat2vol: szvol inconsistent with mat2d nv = 151008, size(mat,2) = 142954 run 2 t= 1.6reading data ... 0.518631Temporally filteringSaving rho1 ERROR: fast_mat2vol: szvol inconsistent with mat2d nv = 151008, size(mat,2) = 142954 Found 0 voxels with corrected AR1 > 0.90Not WhiteningFound 8054 zero-valued voxelsComputing contrasts The res folder in fc.noseed.surf.lh or fc.noseed.surf.rh contains res-001.nii.gz 158KB res-002.nii.gz 158KB that cannot be opended. Thanks Stefano

1 0

Biannual Boston Freesurfer Workshops
by Salar Dini, Elaheh 09 Dec '22

09 Dec '22

External Email - Use Caution Hello dear freesurfer developers, I am leaving this message on the board to ask about the possibility of resuming the Boston Freesurfer course in the near future. The last update that I'm seeing on the Freesurfer official website pertains to the Spring 2020 course which is the one before the pandemic hit. Are you planning on resuming the biannual workshops anytime soon? Thanks and all the best, Ela

1 0

A TRACULA Question
by Salar Dini, Elaheh 09 Dec '22

09 Dec '22

External Email - Use Caution Dear Anastasia, Thanks for your support in the past few months. I got an error message in -prep step: "Exception caught: itk::ExceptionObject (0x8a1e880) Location: "unknown" File: /home1/zs24/sandbox/build/ITKv5-install/include/ITK-5.0/itkSmoothingRecursiveGaussianImageFilter.hxx Line: 229 Description: itk::ERROR: SmoothingRecursiveGaussianImageFilter(0x8a1dd00): The number of pixels along dimension 2 is less than 4. This filter requires a minimum of four pixels along the dimension to be processed. ConvertTransformFile 3 /home1/es2263/TRACULA/subjects_dir/37/dmri/xfms/diff2syn0GenericAffine.mat /home1/es2263/TRACULA/subjects_dir/37/dmri/xfms/diff2syn0GenericAffine.txt --hm --ras file /home1/es2263/TRACULA/subjects_dir/37/dmri/xfms/diff2syn0GenericAffine.mat does not exist . Linux ursapet.med.yale.edu 3.10.0-693.el7.x86_64 #1 SMP Tue Aug 22 21:09:27 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux trac-preproc exited with ERRORS at Thu Oct 20 23:31:04 EDT 2022" I am attaching a log file too. Do you have a guess why this happened and how we can address it? One possibility is that we don't have all the dMRI sequence slices (like 300 out of 1600 are missing). Thanks for your time and help, Ela

1 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