Freesurfer February 2019

freesurfer@nmr.mgh.harvard.edu

114 participants
196 discussions

dt_recon --> tbss
by Daniel Callow 28 Feb '19

28 Feb '19

External Email - Use Caution Hello, I am working through dt_recon troubleshooting (One subjects registration to anatomical is skewed leading to incorrect masking of that subjects FA image). I would like to use the dt_recon's FA outputs and input them into FSL's tbss pipeline. My question is are there any steps I should take converting or modifying the FA images so they are compatible with the tbss pipeline. My intention was to take the dt_recon fa or fa-masked into the tbss_1_preproc command? Are there any considerations I should take into account? Best *Daniel Callow* *PhD Student, Neuroscience and Cognitive Science* Exercise for Brain Health Lab University of Maryland, College Park *ddcc2442(a)gmail.com <ddcc2442(a)gmail.com>* 443-254-6298

2 2

White Matter Edit
by Abazid, Leen 28 Feb '19

28 Feb '19

External Email - Use Caution Dear FreeSurfer supporter, We are working on a PTSD project and we used FreeSurfer to obtain the subcortical and cortical volumes for our MRI images in our patients. We ran our images through the recon-all pipeline. We then extracted the volumes from the lh/rh.aparc.stats files. We plotted the volume data for all the subjects for all the structures using SPSS to compare the volumes between the right and left hemispheres to identify outliers. After further inspection of our plots for the cingulate structure, we believe the volumes are not accurate. The cingulate is very important in the pathogenesis of PTSD and we want to make sure the volumes we are getting are correct. We loaded the data into Freeview to check the gray and white matter segmentations and it appears that the white matter segmentation is including gray matter. We used the following command to edit the wm.mgz: freeview -v mri/brainmask.mgz \ mri/wm.mgz:colormap=heat:opacity=0.4 \ -f surf/lh.white:edgecolor=blue \ surf/lh.pial:edgecolor=red \ surf/rh.white:edgecolor=blue \ surf/rh.pial:edgecolor=red \ surf/rh.inflated:visible=0 \ surf/lh.inflated:visible=0 After editing/cleaning we saved the new wm.mgz image and ran the following command: recon-all -autorecon2-wm -autorecon3 -subjid# Following completion we checked the new lh/rh.aparc.stats files? and the total white matter volume is increasing in some patients and decreasing in others while the gray matter is increasing in some of the subjects (which is what we expect). Do you have any suggestions for why the WM volumes would be increasing if we are removing what we think is not WM? Are we running the correct commands to edit the WM segmentation? Any suggestions or help would greatly appreciated. Thanks, Leen F. Abazid Ph.D. Student Neuroscience Imaging at the Research Imaging Institute UT Health Science Center at San Antonio (210) 823-3532

1 0

mri_segstats and mris_anatomical_stats read out different # of rois
by Winkelbeiner, Stephanie A 28 Feb '19

28 Feb '19

External Email - Use Caution Hi Freesurfers, I'm trying to read out GM and WM values from parcels (ROIs) but get a different amount of ROIs for GM and WM. I use the following commands (embedded in a loop), for GM mris_anatomical_stats -mgz \ -cortex label/lh.cortex.label \ -f stats/lh.250.aparc.pial.stats \ -b -a 250.aparc \ -c label/colortable.ctab \ $subject lh pial and for WM mri_segstats \ --seg mri/vol250.lh.nii.gz \ --ctab ../fsaverage/colortable.ctab \ --i $dtipath/$subj/dti_FA.nii.gz \ --excludeid 0 \ --empty \ --sum stats/${subj}_fa_lh.stats where "../fsaverage/colortable.ctab" has 299 labels and "mri/vol250.lh.nii.gz" is the segmented image that was extended 2mm into the WM and registered to the DTI image. Is it possible that these two commands read out different amounts of ROIs? And is there a reason why the -c flag in mris_anatomical_stats doesn't save the colortable for each subject? I would be very grateful for some help! Thanks, Stephanie The information contained in this electronic e-mail transmission and any attachments are intended only for the use of the individual or entity to whom or to which it is addressed, and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If the reader of this communication is not the intended recipient, or the employee or agent responsible for delivering this communication to the intended recipient, you are hereby notified that any dissemination, distribution, copying or disclosure of this communication and any attachment is strictly prohibited. If you have received this transmission in error, please notify the sender immediately by telephone and electronic mail, and delete the original communication and any attachment from any computer, server or other electronic recording or storage device or medium. Receipt by anyone other than the intended recipient is not a waiver of any attorney-client, physician-patient or other privilege.

2 1

matrix is ill-conditioned or badly scaled, condno = 1e+08
by Arsenije Subotic 28 Feb '19

28 Feb '19

External Email - Use Caution Dear experts, I have recently tried to use the command line to examine thickness differences between two groups after controlling for the effects of age and sex. I get the following error : ERROR: matrix is ill-conditioned or badly scaled, condno = 1e+08 -------------------------------- Possible problem with experimental design: Check for duplicate entries and/or lack of range of continuous variables within a class. If you seek help with this problem, make sure to send: 1. Your command line: mri_glmfit.bin --y CAA_CTRL_RH.thickness.10.mgh --fsgd CAA_CTRL_RH_Age_Sex.fsgd --C Contrasts/CAA-CTRL_Age_Sex.mtx --surf fsaverage rh --cortex --glmdir CAA_CTRL_RH_Age_Sex_Unix.glmdir 2. The FSGD file (if using one) 3. And the design matrix above Attempting to diagnose further SumSq: Min=0.000000 (col 5), Max=611.699463 (col 4) The scale is much different between columns 5 and 4, you may want to normalize by subtracting the mean and dividing by the standard deviation. Column 5, all values are 0 Column 6, all values are 0 Columns 5 and 6 are the same I did not get this error when I was only looking at the difference in thickness between the two groups, so I am guessing there is not an issue with the file format? Looking at the log, does this mean that I also need to demean the age variable? I was planning on running some contrasts looking at the interactions between group and age after controlling for sex etc. in order to see if I should go with a DOSS approach, but using those contrasts still gives me the same error. Thanks for your help! Best, Arsenije --- Arsenije Subotic | MSc Student Department of Clinical Neurosciences Hotchkiss Brain Institute, University of Calgary | T2N 4N1, Canada Tel: (403) 918-6970 arsenije.subotic1(a)ucalgary.ca<mailto:arsenije.subotic1@ucalgary.ca>

2 1

Re: [Freesurfer] downsample fsnative
by Ella Hanns 28 Feb '19

28 Feb '19

2 1

Question about fsaverage_sym
by Alon Baram 28 Feb '19

28 Feb '19

External Email - Use Caution Hello, I have subject's surfaces where I first register the rh to the lh using xhemireg, then register both hemispheres to fsaverage_sym using surfreg, and then resample the registered surface to the isocahedron using mris_surf2surf (see code below if that's helpful, part of the RSA toolbox, written by Joern Diedrichsen). This means that for the two hemispheres of these surfaces, each vertex number corresponds to an exact anatomical location (vertex i in rh will be in the exact mirroring location of vertex i). (I then also move the coordinates of the surface to correspond to world, rather than RAS_tkreg coords, but I think that is irrelevant to my question. This is used to create searchlights ). I then compute curvature maps on these surfaces, and would like to view them in Freeview overlaid on a common surface. As the remapped surfaces were registered to fsaverage_sym, I thought it would make sense to load them as overlays on the fsaverage_sym surfaces. However I'm not sure on which hemisphere of fsaverage_sym should the curv of the rh be overlaid on. I originally thought it shouldn't matter - that the rh of fsaverage_sym was an exact mirror of the lh, so it shouldn't matter if I load the curv as overlay on the lh or rh. But for a reason I don't understand, that is not true, and it really changes how the curvature looks on the surface. Moreover, it looks exactly the same if I load the curvature as overlay on the rh of fsaverage or fsaverage_sym - I thought these two should be different. What am I getting wrong? which is the correct way of viewing the curvature? Thanks very much for the help. ------------------------ function S=freesurfer_mapicosahedron_xhem(subj,subject_dir,varargin); % function S=freesurfer_mapicosahedron(subj,subject_dir,varargin); % Resampels a registered subject surface to a regular isocahedron % This allows things to happen exactly in atlas space - each vertex number % corresponds exactly to a anatomical location % Makes a new folder, called ['x' subj] that contains the remapped subject % Uses function mri_surf2surf % INPUT: % subj: subject name % subjects_dir: freesurfer's SUBJECT_DIR % VARARGIN: % 'hemisphere',[1 2] : left / right or both hemispheres % 'surf_files',{'',''}: Surface files to be resampled % {'.white','.pial','.inflated','.sphere.reg','.sphere'}; % 'curv_files',{'',''}: Curvature files to be resampled % {'.curv','.sulc','.area'}; % 'smoothing',1: Smoothing iterations applied (otherwise nearest neighbour) % --------------------------- % v1.0 Joern Diedrichsen (j.diedrichsen(a)ucl.ac.uk % current_dir=pwd; direct=[getenv('FREESURFER_HOME') filesep 'average' filesep 'surf' ]; old_dir=getenv('SUBJECTS_DIR'); % if (isempty(subject_dir)) % subject_dir=getenv('SUBJECTS_DIR'); % else % setenv('SUBJECTS_DIR',subject_dir); % end; subject_dir=getenv('SUBJECTS_DIR'); structname={'left','right'}; dirname={'LeftHem','RightHem'}; hem={'lh','rh'}; surf_files={'.white','.pial','.inflated'}; curv_files={'.curv','.sulc','.area'}; smoothing=1; hemisphere=[1:2]; % Do both hemispheres vararginoptions(varargin,{'smoothing','surf_files','curv_files','hemisphere' }); % read freesurfer version ver_file = fullfile(getenv('FREESURFER_HOME'),'build-stamp.txt'); if exist(ver_file) fid = fopen(ver_file); verStr = fgetl(fid); fclose(fid); verStr = strrep(verStr,'-v',' '); verStr = textscan(verStr,'%s%f'); fsl_ver = verStr{2}; end new_dir = [subject_dir filesep 'x' subj filesep 'surf']; if (~exist(new_dir)) mkdir(new_dir); end; num_surf=length(surf_files); files={surf_files{:},curv_files{:}}; for h=hemisphere % Original subjects dir if (h==1) orig_dir =fullfile(subject_dir,subj,'surf'); else orig_dir =fullfile(subject_dir,subj,'xhemi','surf'); end; cd(orig_dir); % [vertex_coords, faces] [Vico,F]=read_surf([direct filesep 'lh.sphere.reg']); % /opt/fmrib/FreeSurfer_releases/6.0/average/surf/lh.sphere.reg % Topology file for i=1:length(files) if i<=num_surf % e.g /home/fs0/abaram/scratch/freesurfer-subjects/s99/surf/lh.white [V]=read_surf([orig_dir filesep 'lh' files{i}]); else [V]=read_curv([orig_dir filesep 'lh' files{i}]); end; B=[]; for j=1:size(V,2) % Alon: number of coords - usually 3 % Alon's comment: write the coords of the vertices (one dimension at a % time) as curv files write_curv([orig_dir filesep 'lh.temp.curv'],V(:,j),10000); % e.g [$SUBJECTS_DIR/s99/surf/lh.temp.curv, only x coords of V, number of faces (why only 10000?! I think it's just not used)] if (h==1) system(['DYLD_LIBRARY_PATH=/opt/fmrib/FreeSurfer_releases/6.0/lib/gcc/lib '... 'mri_surf2surf --srcsubject ' subj ' --hemi lh' ... ' --trgsubject ico --trgicoorder 7'... ' --surfreg fsaverage_sym.sphere.reg'... ' --srcsurfval temp.curv --src_type curv' ... ' --trgsurfval tempN.curv --trg_type curv --mapmethod nnf --nsmooth-out ' num2str(smoothing)]); else % Reslice the left hemisphere of the xhemi subject - this % is originally the right hemisphere system(['DYLD_LIBRARY_PATH=/opt/fmrib/FreeSurfer_releases/6.0/lib/gcc/lib '... 'mri_surf2surf --srcsubject ' subj '/xhemi --hemi lh' ... ' --trgsubject ico --trgicoorder 7'... ' --surfreg fsaverage_sym.sphere.reg'... ' --srcsurfval temp.curv --src_type curv' ... ' --trgsurfval tempN.curv --trg_type curv --mapmethod nnf --nsmooth-out ' num2str(smoothing)]); end; % B(:,i)=read_curv([orig_dir filesep hem{h} '.tempN.curv']); % For different alignment, I would have to use -surf_reg option % here % Alon's comment: read in the resampled curv file of a single dim of the % coordinates per vertex. At the end of the j loop, B will have % the all 3 resamped dims of the coordinates. if fsl_ver <= 5.1 B(:,j)=read_curv(fullfile(orig_dir,'tempN.curv')); else B(:,j)=read_curv(fullfile([orig_dir filesep 'lh.tempN.curv' ])); end; end; % write the surfaces (3 col) and curvs (1 col). if i<=num_surf freesurfer_write_surf([new_dir filesep hem{h} files{i}],B(:,1:3),F); else write_curv([new_dir filesep hem{h} files{i}],B(:,1),size(F,1)); end; end; % deleting temporary files try delete(fullfile(orig_dir,['lh.temp.curv'])); if fsl_ver <= 5.1 delete(fullfile(orig_dir,['tempN.curv'])); else delete(fullfile(orig_dir,['lh.tempN.curv'])); end catch err disp(getReport(err)); end; end; setenv('SUBJECTS_DIR',old_dir);

2 1

mni152reg
by john Anderson 28 Feb '19

28 Feb '19

External Email - Use Caution Dear FS experts, I would like to move PET image to MNI152. What is the correct command to do so. I tried mri_vol2vol --mov PET.nii.gz --s subjID --targ $SUBJECTS_DIR/subjID/mri/orig.mga --regheader --o PET_T1.nii.gz mni152reg --s subjID then mri_vol2vol --inv --targ PET_T1.nii.gz --mov $FSL_DIR/data/standard/MNI152_T1_2mm_brain.nii.gz --o PET_T1_2MNI153_2mm.nii.gz --interp nearest --reg $SUBJECTS_DIR/subjID/mri/transforms/reg.mni152.reg.dat I still not happy with thresults. PET not perfectly moved to MNI. please are the steps above correct? Can you kindly suggest any ideas to improve PET registration to MNI? Thanks John

2 4

MD with dt_recon
by Daniel Callow 28 Feb '19

28 Feb '19

External Email - Use Caution Sorry, I just found in a previous post by doug that ADC is the same as MD? *Daniel Callow* *PhD Student, Neuroscience and Cognitive Science* Exercise for Brain Health Lab University of Maryland, College Park *ddcc2442(a)gmail.com <ddcc2442(a)gmail.com>* 443-254-6298

2 2

MD with dt_recon?
by Daniel Callow 28 Feb '19

28 Feb '19

External Email - Use Caution Hello, There is no mean diffusivity output with dt_recon. How would you suggest to create MD images? Could you simply divide the adc.nii.gz map by 3? Best, *Daniel Callow* *PhD Student, Neuroscience and Cognitive Science* Exercise for Brain Health Lab University of Maryland, College Park *ddcc2442(a)gmail.com <ddcc2442(a)gmail.com>* 443-254-6298

2 1

Correcting Defect error
by Miguel Ángel Rivas Fernández 28 Feb '19

28 Feb '19

External Email - Use Caution Dear Freesurfer devs, I ran the recon-all in one subject and and this process was stopped in the topology correction. In particular during the "CORRECTING DEFECT 36 (vertices=19847, convex hull=5892, v0=94733)". I reviewed the lh.inflated.nofix and the lh.orig.nofix files in Freeview and I noticed that the this error would be located in the lh.inflated.nofix file. How can I fix this problem? I sent to you a screenshot of the lh.inflated.nofix file as well as the recon-all.log file if you want take a look. Best regards, -- *Miguel Ángel Rivas Fernández*

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 February 2019