Freesurfer

freesurfer@nmr.mgh.harvard.edu

5 participants
30076 discussions

mri_surfcluster problem
by Emily Cooper 06 Jun '07

06 Jun '07

hi, i am having a strange problem with mri_surfcluster. i give it the following command line: mri_surfcluster --in /disks/gpfs/fmri/emily.permutations/correlation.permutations/perm.mghs/2.lh.ccf.perm.smooth6.mgh --subject fsaverage --hemi lh --thmin 0.5 --thmax 1.0 --sign pos --minarea 6 --sum /disks/gpfs/fmri/emily.permutations/correlation.permutations/perm.clusters/2.ccf.pos.output.sum.txt and i get this result in the summary: # Cluster Growing Summary (mri_surfcluster) # $Id: mri_surfcluster.c,v 1.26.2.4 2006/07/13 17:50:28 greve Exp $ # $Id: mrisurf.c,v 1.441.2.5 2006/11/29 20:28:16 nicks Exp $ # CreationTime 2007/06/01-19:04:15-GMT # cmdline mri_surfcluster --in /disks/gpfs/fmri/emily.permutations/correlation.permutations/perm.mghs/2.lh.ccf.perm.smooth6.mgh --subject fsaverage --hemi lh --thmin 0.5 --thmax 1.0 --sign pos --minarea 6 --sum /disks/gpfs/fmri/emily.permutations/correlation.permutations/perm.clusters/2.ccf.pos.output.sum.txt # cwd /disks/gpfs/fmri/emily.permutations/correlation.permutations # sysname Linux # hostname andrew.bsd.uchicago.edu # machine i686 # FixVertexAreaFlag = 1 # # Input /disks/gpfs/fmri/emily.permutations/correlation.permutations/perm.mghs/2.lh.ccf.perm.smooth6.mgh # Frame Number 0 # srcsubj fsaverage # hemi lh # surface white # SUBJECTS_DIR /home/ecooper/subjects # Minimum Threshold 0.5 # Maximum Threshold 1 # Threshold Sign pos # AdjustThreshWhenOneTail 1 # Area Threshold 6 mm^2 # Overall max 0.632052 at vertex 5494 # Overall min 0.12689 at vertex 750 # NClusters 2 # Total Cortical Surface Area 70394.7 (mm^2) # FixMNI = 1 # # ClusterNo Max VtxMax Size(mm^2) TalX TalY TalZ NVtxs 1 0.632 5494 85426.29 -50.2 -16.0 -7.2 161181 2 0.223 9300 23.61 -31.7 -26.7 9.9 68 the problem is that this is just entirely incorrect. when i view the input dataset in tksurfer, it's clear just from scanning over the vertices that 161181 out of 163842 total of them do not have intensity values between .5 and 1 (thmin and thmx), let alone in a giant cluster together. furthermore, when i make a historgram of a text file containing these intensity values, it shows that the majority of them are less than .5, and thus shouldn't be recognized as part of a cluster based on my specifications in the mri_surfcluster command: $breaks [1] -0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 [13] 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 $counts [1] 10 205 1320 3881 7834 12333 16777 20168 21369 20405 17219 13789 [13] 10453 7479 4897 2988 1599 737 283 81 14 1 $intensities [1] 0.0012206880 0.0250241086 0.1611308456 0.4737490997 0.9562871547 [6] 1.5054747867 2.0479486334 2.4618840102 2.6084886659 2.4908143211 [11] 2.1019030529 1.6832069921 1.2759854006 0.9129527227 0.5977710233 [16] 0.3647416413 0.1951880470 0.0899647221 0.0345454767 0.0098875746 [21] 0.0017089635 0.0001220688 $density [1] 0.0012206880 0.0250241086 0.1611308456 0.4737490997 0.9562871547 [6] 1.5054747867 2.0479486334 2.4618840102 2.6084886659 2.4908143211 [11] 2.1019030529 1.6832069921 1.2759854006 0.9129527227 0.5977710233 [16] 0.3647416413 0.1951880470 0.0899647221 0.0345454767 0.0098875746 [21] 0.0017089635 0.0001220688 $mids [1] -0.075 -0.025 0.025 0.075 0.125 0.175 0.225 0.275 0.325 0.375 [11] 0.425 0.475 0.525 0.575 0.625 0.675 0.725 0.775 0.825 0.875 [21] 0.925 0.975 any ideas as to what is going wrong here? am i using the thmin and thmax flags wrong? thanks so much, emily

2 3

subcortical segmentation
by Mishkin Derakhshan 05 Jun '07

05 Jun '07

Hi, I am interested in just obtaining the subcortical segmentation labels and not generating surfaces. I would like to do this in as little time as possible and since I am only looking for label volumes (ie. hippocampal volumes) i figured that there must be some steps I can leave out of -autorecon2. Can anyone tell me the flags I should set to get the fastest way to generate these label files? I found: autorecon-all -subcortseg -subjid <subject> on the wiki, but it says this may take 11+ hours to complete so I thought I would just check before I batched things to run. many thanks, mishkin

5 9

scaling thickness values
by Goldman, Aaron (NIH/NIMH) [C] 05 Jun '07

05 Jun '07

Hi, regarding thickness analysis using Freesurfer data, I was wondering if it was recommended to use any kind of scaling (for example, to the subjects' overall average thickness). I'm interested regarding both node-by-node analysis using mris_glmfit, and region-based using the aparc.stats files. Thanks, -Aaron-

2 1

Re: [Freesurfer] fmri conjunction/correlation map
by Cameron Ellis 05 Jun '07

05 Jun '07

Hi all, I have a few questions on two related topics. Topic 1, conjunction analysis: Thank you Doug for your response pointing out the mergecontrasts tool. I have 2 questions for anyone who may have used this in the past. First, do I specify the two contrasts to merge using the contrast flag twice? Ex: -contrast [name] [thresh] [tail] -contrast [name] [thresh] [tail] or do I specify both contrasts under a single tag? Additionally, this is a memory task and the encoding and retrieval runs were analyzed separately (same analysis type, but same coding in paradigm files and different sessions parent directories) so will it be possible to indicate to the program where the encoding contrast to be merged with the retrieval contrast is? Or will I have to rearrange my directories, recode my paradigm files, and make new contrasts? probably the latter huh? Topic 2, correlation analysis: As an alternative to a conjunction analysis, is it possible to use freesurfer to perform a voxel-by-voxel group level correlation analysis of encoding versus retrieval runs for a memory task perhaps obtaining a map of r-values or sigs or r values representing the correlation of the time courses from corresponding voxels from the two runs? what freesurfer tools would do this? Thanks in advance, Cameron > Oh, you want a "conjunction" I think. Corrlation means something else. > Try mergecontrasts-sess > > doug > > > > Cameron Ellis wrote: > >>Hi Doug, >>We would like to compare, voxel by voxel, the activity between two >>back-to-back runs, one encoding run and one retrieval run for a memory >>task. Ideally what we have in mind would be to view, on the inflated >>surface, a map of voxels that were above a certain threshold during both >>encoding and retrieval. Could we somehow make a whole brain functional >>mask from the encoding run an apply this to the retrieval run? >>Cameron >> >> >> >>>I'm not sure what you are trying to do. sound like you have two values >>>per voxel for each subject that you want to correlate, but you can't >>>really correlate two values. I think I'm missing something ... >>> >>>doug >>> >>>Cameron Ellis wrote: >>> >>> >>> >>>>Does anyone have an answer to this question? >>>> >>>>Hello, >>>> >>>>I am looking for a way to perform voxel-wise, within-subject >>>> correlations >>>>of two >>>>activation maps (in talairach space). The two existing maps reflect >>>>encoding >>>>and retrieval runs of a memory task. The idea would be to generate a >>>>correlation map across all subjects, to identify clusters where there >>>> was >>>>a >>>>positive or negative correlation between encoding and retrieval >>>>activation. >>>> >>>>I've seen this done with SPM but not Freesurfer. Any suggestions? >>>>(Perhaps >>>>this would involve glmfit?) >>>> >>>>Many thanks, >>>>Josh >>>>_______________________________________________ >>>>Freesurfer mailing list >>>>Freesurfer(a)nmr.mgh.harvard.edu >>>>https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>> >>>> >>>> >>>> >>>> >>>> >>>-- >>>Douglas N. Greve, Ph.D. >>>MGH-NMR Center >>>greve(a)nmr.mgh.harvard.edu >>>Phone Number: 617-724-2358 >>>Fax: 617-726-7422 >>> >>>In order to help us help you, please follow the steps in: >>>surfer.nmr.mgh.harvard.edu/fswiki/BugReporting >>> >>> >>> >>> >>> >>> >> >> >> >> > > -- > Douglas N. Greve, Ph.D. > MGH-NMR Center > greve(a)nmr.mgh.harvard.edu > Phone Number: 617-724-2358 > Fax: 617-726-7422 > > In order to help us help you, please follow the steps in: > surfer.nmr.mgh.harvard.edu/fswiki/BugReporting > > >

2 1

tkmedit-sess question
by Avram Holmes 05 Jun '07

05 Jun '07

FreeSurfer Folks, I am attempting to look at some data with tkmedit-sess tkmedit-sess -analysis LowTraumaSPMHRF -contrast all_v_fix -d . -s FPPREW09 -map sig -fsd bold but I am getting the following error. corRead(): can't open file /exp/REfMRI/fMRI_FPPREW/structural//FPPREW09/mri/orig/COR-.info Error: Loading volume orig Couldn't read the anatomical volume. Tkmedit couldn't read the volume you specified. This could be because the image format wasn't recognized, or it couldn't find the proper header, or the file(s) were unreadable, or it was the wrong size. I can look at the data with tkmedit (without the sess) loading all the files manually. I have tried the tkmedit-sess several different ways but I always get an error. Does anyone have an idea about how I get this to work? Thanks, Avram ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Avram Holmes Department of Psychology Harvard University 1210 William James Hall Phone:(617) 495-0790 33 Kirkland Street Fax:(617) 495-3728 Cambridge, MA 02138, USA Email: holmes(a)fas.harvard.edu ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

2 4

how to import private color scales
by Istvan Akos Morocz 04 Jun '07

04 Jun '07

dear freesurfers, how do i import my own 'private' colorscales, i.e., colormaps for the overlay data ? something richer in colors than 'heat' ? what's the format and size for such a file? thanks, bye, pisti ----- pisti(a)bwh.harvard.edu

2 1

problem with tksurfer (fwd)
by Claus Tempelmann 04 Jun '07

04 Jun '07

Hello everybody, Having worked with quite old versions of freesurfer for a long time we decided recently to install a newer version on two new computers. Both are running under the 64 bit version of SUSE Linux 10.2, both have a Dual Opteron processor, they have 6 GB (system 1) or 8 GB (system 2) and most hardware and software is quite comparable in my opinion. There is one major difference with the graphic cards, system 1 has an MSI GEForce 6600, whereas system 2 has an GEForce FX 7600GS. I'd expect that both cards are more than sufficient for freesurfer and I doubt that this difference can be the reason for the difference I have observed but may be some of you have experienced a similar problem and can help me out. What's the problem? All analysis steps of the standard procedure seem to work fine on both systems. However, I see a reproducible strange problem on system 2 when I try to use tksurfer on the standard data set (bert) coming with the package (freesurfer-Linux-rh9-stable- pub-v3.0.5.-full.tar). When I run the command tkregister bert lh inflated on system 1, everything looks like expected. But on system 2 something is wrong, the program tells the number of vertices and faces as usual and the numbers are the same as on system 2 (as they should, I'm using the same data set). Nevertheless, the output looks quite different, I'll try to attach a screenshot. The window shows up as usual and a small part of the inflated brain is shown at the right edge of the window but the rest of the brain is never shown. When I rotate this surface the visible part of the inflated brain is rotated correctly but the rest is always missing. Can anybody help? Thanks in advance Claus ----------------------------------------------------------------------- Claus Tempelmann email: Claus.Tempelmann(a)medizin.uni-magdeburg.de Clinic for Neurology II MR department Tel.: +49-(0)391-6117183 OvG University Magdeburg +49-(0)391-6117177 Leipziger Strasse 44 39120 Magdeburg Fax: +49-(0)391-6117178 Germany

2 2

Masking at the individual level
by Dan Dillon 04 Jun '07

04 Jun '07

Dear FreeSurfers, Quick question about masking. I routinely make a brainmask for each subject and include the mask flag in the mkanalysis (see below), but I note that when I run contrasts and plot results-either for individual subjects or for group-level stuff-I get statistical results for voxels outside the brain. I'd like to mask the extra-brain voxels out at the individual level, so that they don't get included at any point in my group level analyses. How do I do this? More generally, does the brainmask serve only a visual/display function or can I use it to constrain what I'm doing stats on (which is what I want it to do)? Thanks for your endless help! Dan Dillon PS. This is a a typical mkanalysis step . . . mkanalysis-sess.new -analysis ${analysis} -gammafit 2.25 1.25 -TR 2.5 -paradigm mg_3_7.par -designtype event-related -funcstem fmcs tcsm6 -nconditions 13 -polyfit 2 -taumax 30 -inorm -mask brain -timeoffset -1.25 -TER .1 -autostimdur -force

2 1

ANCOVA model in Freesurfer
by Juergen Haenggi 04 Jun '07

04 Jun '07

Dear Freesurfer users I am trying to create the GroupDescriptorFile 1 for an ANCOVA in order to analyze local cortical thickness. I intend to compare two groups while controlling for age and mean cortical thickness. Am I right that I can leave out the term "class" and model group as a variable? GroupDescriptorFile 1 Title CorticalThickness Variables Group Age MeanThickness Input Control01 ControlSubject 27 3.14 Input Control02 ControlSubject 31 3.28 .... Does the variables all have to be integers or can it be also a string? Thanks for helping Best regards Juergen ----------------------------------------------------------- Juergen Haenggi Ph.D. (Dr. des.) Division of Neuropsychology Institute of Psychology University of Zurich Binzmuehlestrasse 14, PO Box 25 8050 Zurich, Switzerland 0041 44 635 73 97 (phone office) 0041 76 445 86 84 (phone mobile) 0041 44 635 74 09 (fax office) BIN 4.D.04 (office room number) j.haenggi(a)psychologie.uzh.ch (email) http://www.psychologie.uzh.ch/neuropsy/ (website) http://www.juergenhaenggi.ch (private website) -----------------------------------------------------------

2 1

Situation with OASIS
by Martin Kavec 04 Jun '07

04 Jun '07

Hi, I was wondaring, what is the situation with OASIS data? Could you please share with us, in which form you will make the results of your great effort available? Thanks in advance for the great job (and reply of course). Martin

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