Freesurfer

freesurfer@nmr.mgh.harvard.edu

19 participants
30132 discussions

Aparc Stats & Different versions of Freesurfer
by Sherri Novis 14 Jul '09

14 Jul '09

Hello - I am comparing aparc stats across several groups of data, and just noticed that recently processed groups do not have corpus callosum values... which made me wonder if different versions of freesurfer will produce different aparc data... Can aparc stats from an older version of freesurfer be compared to aparc stats from a more recent version? Or do they all have to be processed with the same version? Also - we noticed that the aparc areas do not extract in the same order... is this also because of the different version used to process the data? Thanks much, Sherri

2 1

co-registration problem between single subject and fsaverage space
by leemon albert 14 Jul '09

14 Jul '09

Dear FreeSurfer Community... 1. For QDEC analysis, we have pre-computed smoothed data for the target fsaverage surface using the command: recon-all -s <subjid> -qcache. then we obtained 'lh.thickness.fsaverage.asc ' using mris_convert. The procedure is here mris_convert -c ./lh.thickness.fsaverage.mgh \ $SUBJECTS_DIR/fsaverage/surf/lh.white \ lh.thickness.fsaverage.asc Finally, we read the 'lh.thickness.fsaverage.asc ' in matlab using read_asc.m and met the following error. [S, v, f] = read_asc('/usr/local/freesurfer/subjects/lh.thickness.fsaverage.asc'); ??? Error using ==> fscanf Invalid size. Error in ==> read_asc at 45 f = fscanf(fp, '%d', S(3)*5); How to fix this problem? 2. In fact, I want to use BA Labels ( http://surfer.nmr.mgh.harvard.edu/fswiki/BrodmannAreaMaps) to extract the thickness of visual area. And am I right in assuming that the contents of this file, are my thickness stats of my subject after being aligned to fsaverage space, and that each line is in the correct order that corresponds to each vertex in fsaverage/surf/lh.{pial,white} ? 3. What is the best approach to co-register a single subject to fsaverage space and for the results to have identical number of vertices and faces as fsaverage? Any help would be appreciated. Thanks -- Albert Lab of Complex System and Artificial Intelligence, Institute of Automation,Chinese Academy of Sciences(CASIA) Beijing, 100190,P.R.CHINA

1 0

asegstats2tabel, ICV
by Rosa Steimke 14 Jul '09

14 Jul '09

HI freesurfer experts! i want to correct for intercranial volume in my qdec analysis. To do that I want to generate a file in which the intracranial volume of each subject is listed. do i have to use the asegstats2tabel command? I tried it and it gave me following error message: asegstats2table --subjects 1 2 4 5 6 7 8 9 10 11 12 13 14 15 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 41 42 43 45 47 48 --meas volume --tablefile aseg_stats.txt Traceback (most recent call last): File "/Applications/freesurfer/bin/asegstats2table", line 6, in <module> import fsutils ImportError: No module named fsutils [admins-imac:~/Rosa_MRTs] steimke% asegstats2table --subjects 1 2 4 5 6 7 8 9 10 11 12 13 14 15 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 41 42 43 45 47 48 Traceback (most recent call last): File "/Applications/freesurfer/bin/asegstats2table", line 6, in <module> import fsutils ImportError: No module named fsutils what did i do wrong? is this the right approach to get a ICV file? thanks in advance!!! Rosa Rosa Steimke Student Master of Neuroscience Vrije Universiteit Amsterdam Netherlands -- Neu: GMX Doppel-FLAT mit Internet-Flatrate + Telefon-Flatrate für nur 19,99 Euro/mtl.!* http://portal.gmx.net/de/go/dsl02

3 2

autorecon2-error
by whblqh＠mail.ustc.edu.cn 14 Jul '09

14 Jul '09

Hello,freesurfer experts I test freesurfer on my computer and command "recon-all -s bert -autorecon2",but there were some errors,as follows: > ......... > #-------------------------------------------- > #@# Cortical ribbon mask Mon Jul 13 02:59:19 CST 2009 > /home/whb/soft/freesurfer/subjects/bert/mri > > mris_volmask --label_left_white 2 --label_left_ribbon 3 --label_right_white 41 > --label_right_ribbon 42 --save_ribbon --save_distance bert > > mris_volmask.bin: error while loading shared libraries: > /home/whb/soft/freesurfer/lib/vtk/lib/vtk-5.2/libvtkRendering.so.5.2: cannot > restore segment prot after reloc: Permission denied > Linux localhost.localdomain 2.6.21-1.3194.fc7 #1 SMP Wed May 23 22:35:01 EDT 2007 > i686 i686 i386 GNU/Linux > > recon-all exited with ERRORS at Mon Jul 13 02:59:20 CST 2009 > > I don't know what's the reason. would you like to help me£¿ > > Thanks in advance. > > > Hai-bao Wang > > Jul-14, 2009 > > >

1 0

Re: [Freesurfer] mri_convert
by Feng-Xian 14 Jul '09

14 Jul '09

Hi, Thank you for your response. I want to make sure the command as follows. If I want to obtain the thickness ascii with Tal coordinate, can the following commands make I to obtain? First, I change the coordinate from Freesurfer to Tal using the command. mris_convert –t subjid lh.pial lh.pial.tal Then, I obtain the thickness ascii with Tal coordinate from the following command. mris_convert -c lh.thickness lh.pial.tal lh.thickness.asc Finally, I want to check whether the coordinate is true or not. So I open tksurfer to check. But, I found the vertex Tal coordinate of lh.thickness.asc is different from the shows of terminal. Have something wrong of these commands? Thank you in advance! Feng-Xian On Mon, 19 Jan 2009 10:01:21 -0500 (EST), Bruce Fischl wrote > read the mris_convert help. You can convert the surfaces to ascii as > well > (e.g. mris_convert lh.white lh.white.asc) > > cheers > Bruce > On Mon, 19 Jan 2009, Feng-Xian > wrote: > > > > > Hi Bruce > > Thanks for your answer. But I really want to get the result of surfaces > > on FreeSurfer and I would like to read these surfaces (e.g. lh inflated, lh > > thickness) on Windows XP. > > > > Would you teach me how can I get these? > > > > Thanks in advance! > > Feng-Xian > > > > > > > > On Mon, 19 Jan 2009 09:05:59 -0500 (EST), Bruce Fischl wrote > >> Hi Feng-Xian, > >> > >> mri_convert only works on volumes, and to convert e.g. the orig.mgz > >> volume to nifti use: > >> > >> mri_convert orig.mgz orig.nii > >> > >> cheers, > >> Bruce > >> > >> On Mon, 19 > >> Jan 2009, Feng-Xian wrote: > >> > >>> > >>> Hi, > >>> I am finished to run my subjects using recon-all and I would like to > > know > >>> the usage of mri_convert to convert the results of FreeSurfer into the > >>> analyze format or nifti. > >>> > >>> Some of the results of FreeSurfer are volumes (e.g. brainmask.mgz) and > > some > >>> of them are surfaces (e.g. lh inflated). > >>> > >>> What can I do to convert these formats to analysis format or nifti? > >>> > >>> Thanks in advance! > >>> Feng-Xian > >>> > >>> ------------------------------------------------- > >>> Chang Gung University Mail Server Web Mail system > >>> By Chang-Gung University Information Center > >>> (http://www.cgu.edu.tw/ic/) > >>> > >>> _______________________________________________ > >>> Freesurfer mailing list > >>> Freesurfer(a)nmr.mgh.harvard.edu > >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer > >>> > >>> > >>> > > > > > > ------------------------------------------------- > > Chang Gung University Mail Server Web Mail system > > By Chang-Gung University Information Center > > (http://www.cgu.edu.tw/ic/) > > > > > > ------------------------------------------------- Chang Gung University Mail Server Web Mail system By Chang-Gung University Information Center (http://www.cgu.edu.tw/ic/)

1 0

Fornix Creating Surface
by True Price 13 Jul '09

13 Jul '09

So we're having a problem with the segmentation of the fornix. It is currently creating a cortical surface (i.e. a separation between the main and pial surfaces) despite removing the fornix from the brainmask. We have also tried filling in the white matter mask without any improvement. Do you have any suggestions on how to fix this? We have included a couple of pictures for your reference- the brainmask (with the deleted fornix) with the main and pial surfaces visible, and the corresponding white matter mask. Thanks for your response, True Price Summer Intern- King Lab

2 1

recon-all taking long time
by Barnali Basu 13 Jul '09

13 Jul '09

Experts, I have been working on Freesurfer last 5 months and initially the recon processes were within normal time limits. However, I am working on a different set of subjects now and its consuming days together. For my current data,autorecon1 is fine, it stops at a particular point in autorecon2 for infinity. Any suggestions would be helpful. I am pasting the autorecon2 process here: recon-all -autorecon2 -subjid ja Subject Stamp: freesurfer-Linux-centos4-stable-pub-v4.1.0 Current Stamp: freesurfer-Linux-centos4-stable-pub-v4.1.0 INFO: SUBJECTS_DIR is /usr/local/freesurfer/subjects Actual FREESURFER_HOME /usr/local/freesurfer -rw-rw-r-- 1 barnali barnali 257784 Jul 9 10:32 /usr/local/freesurfer/subjects/ja/scripts/recon-all.log Linux barnali-work 2.6.24-16-server #1 SMP Thu Apr 10 13:58:00 UTC 2008 i686 GNU/Linux #------------------------------------- #@# EM Registration Thu Jul 9 10:35:21 PDT 2009 /usr/local/freesurfer/subjects/ja/mri mri_em_register -mask brainmask.mgz nu.mgz /usr/local/freesurfer/average/RB_all_2008-03-26.gca transforms/talairach.lta using MR volume brainmask.mgz to mask input volume... reading 1 input volumes... logging results to talairach.log reading '/usr/local/freesurfer/average/RB_all_2008-03-26.gca'... average std = 6.9 using min determinant for regularization = 4.7 0 singular and 1812 ill-conditioned covariance matrices regularized reading 'nu.mgz'... freeing gibbs priors...done. bounding unknown intensity as < 14.9 or > 790.2 total sample mean = 84.0 (478 zeros) ************************************************ spacing=8, using 2185 sample points, tol=1.00e-05... ************************************************ register_mri: find_optimal_transform find_optimal_transform: nsamples 2185, passno 0, spacing 8 resetting wm mean[0]: 102 --> 107 resetting gm mean[0]: 64 --> 64 input volume #1 is the most T1-like using real data threshold=8.0 using (107, 91, 112) as brain centroid... mean wm in atlas = 107, using box (90,77,91) --> (123, 104,132) to find MRI wm before smoothing, mri peak at 137 after smoothing, mri peak at 151, scaling input intensities by 0.709 scaling channel 0 by 0.708609 initial log_p = -92132.8 ************************************************ First Search limited to translation only. ************************************************ Found translation: (-2.8, -1.7, 1.7): log p = -39503.6 **************************************** Nine parameter search. iteration 0 nscales = 0 ... **************************************** Result so far: scale 1.000: max_log_p=-37619.9, old_max_log_p =-39503.6 (thresh=-39464.1) 1.062 0.000 0.000 -10.944; 0.000 1.053 0.139 -22.515; 0.000 -0.131 0.991 18.328; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 1 nscales = 0 ... **************************************** Result so far: scale 1.000: max_log_p=-35219.5, old_max_log_p =-37619.9 (thresh=-37582.3) 1.062 0.000 0.000 -10.944; 0.000 1.266 0.027 -29.117; 0.000 0.025 1.003 -1.688; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 2 nscales = 0 ... **************************************** Result so far: scale 1.000: max_log_p=-35219.5, old_max_log_p =-35219.5 (thresh=-35184.3) 1.062 0.000 0.000 -10.944; 0.000 1.266 0.027 -29.117; 0.000 0.025 1.003 -1.688; 0.000 0.000 0.000 1.000; reducing scale to 0.2500 **************************************** Nine parameter search. iteration 3 nscales = 1 ... **************************************** Result so far: scale 0.250: max_log_p=-32338.2, old_max_log_p =-35219.5 (thresh=-35184.3) 1.096 0.000 0.000 -15.249; 0.000 1.208 0.026 -25.944; 0.000 0.026 1.035 -8.782; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 4 nscales = 1 ... **************************************** Result so far: scale 0.250: max_log_p=-31970.6, old_max_log_p =-32338.2 (thresh=-32305.8) 1.096 0.000 0.000 -17.124; 0.000 1.189 0.025 -21.806; 0.000 0.026 1.035 -6.907; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 5 nscales = 1 ... **************************************** Result so far: scale 0.250: max_log_p=-31970.6, old_max_log_p =-31970.6 (thresh=-31938.6) 1.096 0.000 0.000 -17.124; 0.000 1.189 0.025 -21.806; 0.000 0.026 1.035 -6.907; 0.000 0.000 0.000 1.000; reducing scale to 0.0625 **************************************** Nine parameter search. iteration 6 nscales = 2 ... **************************************** Result so far: scale 0.062: max_log_p=-31028.4, old_max_log_p =-31970.6 (thresh=-31938.6) 1.083 0.000 0.000 -14.039; 0.000 1.194 0.025 -22.828; 0.000 0.026 1.039 -7.345; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 7 nscales = 2 ... **************************************** Result so far: scale 0.062: max_log_p=-30899.3, old_max_log_p =-31028.4 (thresh=-30997.4) 1.079 0.000 0.000 -13.957; 0.000 1.194 0.025 -22.828; 0.000 0.026 1.043 -7.785; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 8 nscales = 2 ... **************************************** Result so far: scale 0.062: max_log_p=-30683.0, old_max_log_p =-30899.3 (thresh=-30868.4) 1.070 0.000 0.000 -12.858; 0.000 1.194 0.025 -22.828; 0.000 0.026 1.039 -7.344; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 9 nscales = 2 ... **************************************** Result so far: scale 0.062: max_log_p=-30108.0, old_max_log_p =-30683.0 (thresh=-30652.3) 1.058 0.000 0.000 -11.225; 0.000 1.194 0.025 -22.828; 0.000 0.026 1.051 -9.140; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 10 nscales = 2 ... **************************************** Result so far: scale 0.062: max_log_p=-29914.1, old_max_log_p =-30108.0 (thresh=-30077.9) 1.058 0.000 0.000 -11.225; 0.000 1.180 0.025 -21.172; 0.000 0.027 1.059 -10.034; 0.000 0.000 0.000 1.000; **************************************** Nine parameter search. iteration 11 nscales = 2 ... **************************************** Result so far: scale 0.062: max_log_p=-29914.1, old_max_log_p =-29914.1 (thresh=-29884.1) 1.058 0.000 0.000 -11.225; 0.000 1.180 0.025 -21.172; 0.000 0.027 1.059 -10.034; 0.000 0.000 0.000 1.000; min search scale 0.025000 reached *********************************************** Computing MAP estimate using 2185 samples... *********************************************** dt = 5.00e-06, momentum=0.80, tol=1.00e-05 l_intensity = 1.0000 Aligning input volume to GCA... Transform matrix 1.05773 0.00000 0.00000 -11.22528; 0.00000 1.17956 0.02501 -21.17203; 0.00000 0.02668 1.05912 -10.03400; 0.00000 0.00000 0.00000 1.00000; nsamples 2185 Quasinewton: input matrix 1.05773 0.00000 0.00000 -11.22528; 0.00000 1.17956 0.02501 -21.17203; 0.00000 0.02668 1.05912 -10.03400; 0.00000 0.00000 0.00000 1.00000; v3p/netlib/opt/lbfgs.c: lb3_1.lp > 0 outof QuasiNewtonEMA: 012: -log(p) = 29914.1 tol 0.000010 Resulting transform: 1.058 0.000 0.000 -11.225; 0.000 1.180 0.025 -21.172; 0.000 0.027 1.059 -10.034; 0.000 0.000 0.000 1.000; pass 1, spacing 8: log(p) = -29914.1 (old=-92132.8) transform before final EM align: 1.058 0.000 0.000 -11.225; 0.000 1.180 0.025 -21.172; 0.000 0.027 1.059 -10.034; 0.000 0.000 0.000 1.000; ************************************************** EM alignment process ... Computing final MAP estimate using 244171 samples. ************************************************** dt = 5.00e-06, momentum=0.80, tol=1.00e-07 l_intensity = 1.0000 Aligning input volume to GCA... Transform matrix 1.05773 0.00000 0.00000 -11.22528; 0.00000 1.17956 0.02501 -21.17203; 0.00000 0.02668 1.05912 -10.03400; 0.00000 0.00000 0.00000 1.00000; nsamples 244171 Quasinewton: input matrix 1.05773 0.00000 0.00000 -11.22528; 0.00000 1.17956 0.02501 -21.17203; 0.00000 0.02668 1.05912 -10.03400; 0.00000 0.00000 0.00000 1.00000; dfp_em_step_func: 011: -log(p) = 3772889.8 after pass:transform: ( 1.07, 0.01, 0.01, -11.23) ( 0.00, 1.18, 0.03, -21.17) ( 0.00, 0.03, 1.06, -10.03) v3p/netlib/opt/lbfgs.c: lb3_1.lp > 0 pass 2 through quasi-newton minimization... v3p/netlib/opt/lbfgs.c: lb3_1.lp > 0 outof QuasiNewtonEMA: 013: -log(p) = 3772889.8 tol 0.000000 final transform: 1.067 0.008 0.009 -11.225; 0.001 1.180 0.027 -21.172; 0.001 0.027 1.059 -10.034; 0.000 0.000 0.000 1.000; writing output transformation to transforms/talairach.lta... registration took 33 minutes and 19 seconds. #-------------------------------------- #@# CA Normalize Thu Jul 9 11:08:40 PDT 2009 /usr/local/freesurfer/subjects/ja/mri mri_ca_normalize -mask brainmask.mgz nu.mgz /usr/local/freesurfer/average/RB_all_2008-03-26.gca transforms/talairach.lta norm.mgz using MR volume brainmask.mgz to mask input volume... reading 1 input volumes reading atlas from '/usr/local/freesurfer/average/RB_all_2008-03-26.gca'... reading transform from 'transforms/talairach.lta'... reading input volume from nu.mgz... resetting wm mean[0]: 102 --> 107 resetting gm mean[0]: 64 --> 64 input volume #1 is the most T1-like using real data threshold=8.0 using (107, 91, 112) as brain centroid... mean wm in atlas = 107, using box (90,77,91) --> (123, 104,132) to find MRI wm before smoothing, mri peak at 137 after smoothing, mri peak at 151, scaling input intensities by 0.709 scaling channel 0 by 0.708609 using 244171 sample points... INFO: compute sample coordinates transform 1.067 0.008 0.009 -11.225; 0.001 1.180 0.027 -21.172; 0.001 0.027 1.059 -10.034; 0.000 0.000 0.000 1.000; INFO: transform used finding control points in Left_Cerebral_White_Matter.... found 41584 control points for structure... bounding box (125, 59, 30) --> (189, 159, 190) It stops here and doesnt process any further. I tried twice, with same result. I work on a dual core Ubuntu 5GB machine Thanks Barnali

2 3

Fornix Creating Surface
by True Price 13 Jul '09

13 Jul '09

1 0

Sobel Filter
by jentis＠nmr.mgh.harvard.edu 13 Jul '09

13 Jul '09

Hi Team, Does Freeview have a gradient filter (such as a sobel filter) or any kind of operator? Thanks! Best, Jon

1 0

graymid surfaces
by Jakob Heinzle 13 Jul '09

13 Jul '09

Dear freesurfer experts, I would like to create (in addition to the pial and white surfaces) the middle between the two. I know that I can do it by using the -graymid options. There are 2 questions I have with respect to that option. How can I do it without having to recalculate the other surfaces, which seems to happen, if I run e.g. " mris_make_surfaces -graymid -MGZ subject_id lh/rh"? Is there a away of telling recon-all to directly calculate the graymid surface? Thanks, Jakob -- ---------------------------------- Jakob Heinzle, Dr. sc. nat. (PhD) Bernstein Center for Computational Neuroscience Philippstrasse 13 Haus 6 D - 10115 Berlin e-mail: jakob.heinzle(a)bccn-berlin.de Phone: +49 (0) 30 2093 6767 Fax: +49 (0) 30 2093 6771 ----------------------------------

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