Freesurfer November 2014

freesurfer@nmr.mgh.harvard.edu

99 participants
144 discussions

error with fast_selxavg3
by Francesca Strappini 10 Nov '14

10 Nov '14

Dear all, I'm trying to run a retinotopy analysis in several subjects but it crashed with an error in one subject. I ran this command line: selxavg3-sess -a rtopy.HS.rh -s SUBJ09 and this is the last part of the output with the error message. Any help would be much appreciated. Best Francesca < M A T L A B (R) > Copyright 1984-2013 The MathWorks, Inc. R2013b (8.2.0.701) 64-bit (glnxa64) August 13, 2013 ---------------------------------------------------- Your MATLAB license will expire in 56 days. Please contact your system administrator or MathWorks to renew this license. ---------------------------------------------------- To get started, type one of these: helpwin, helpdesk, or demo. For product information, visit www.mathworks.com. >> >> >> >> >> >> >> sxa3pwd = /usr/local/freesurfer/fsfast/MeytalRetinotopySmoothed >> sxa3cmd = /usr/local/freesurfer/fsfast/bin/selxavg3-sess -a rtopy.HS.rh -s SUBJ09 >> /usr/local/freesurfer/fsfast/toolbox/fast_selxavg3.m >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> >> #@# SUBJ09 ############################### /usr/local/freesurfer/fsfast/MeytalRetinotopySmoothed/SUBJ09 ------------------------- $Id: fast_selxavg3.m,v 1.100.2.2 2012/11/30 18:40:38 greve Exp $ /usr/local/freesurfer/fsfast/toolbox/fast_selxavg3.m /usr/local/freesurfer/fsfast/toolbox/fast_ldanaflac.m /usr/local/freesurfer/matlab/MRIread.m ------------------------- outtop = /usr/local/freesurfer/fsfast/MeytalRetinotopySmoothed Extension format = nii.gz nruns = 4 autostimdur = outanadir = /usr/local/freesurfer/fsfast/MeytalRetinotopySmoothed/SUBJ09/bold/rtopy.HS.rh Found 110059/126364 (87.1) voxels in mask Creating Design Matrix ... creation time = 0.005 sec DoMCFit = 1 ntptot = 960, nX = 36, DOF = 924 Saving X matrix to /usr/local/freesurfer/fsfast/MeytalRetinotopySmoothed/SUBJ09/bold/rtopy.HS.rh/Xtmp.mat Error using svd Input to SVD must not contain NaN or Inf. Error in cond (line 36) s = svd(A); Error in fast_selxavg3 (line 279) XCond = cond(XtX); -- Francesca Strappini, Ph.D. Neurobiology Department Weizmann Institute of Science 234 Herzl Street, Rehovot 7610001 Israel Tel.: +972 58 444 2584 E-mail: francesca.strappini(a)weizmann.ac.il

2 5

edit failure
by Octavian Lie 10 Nov '14

10 Nov '14

Dear all, 1. in the attached there is a gyrus that does not incorporated into surfaces, whatever I do (wm edits/topological defects-cloning voxels from brainmask.mgz or T1.mgz into wm.mgz, save volume then run recon-all autorecon2-wm autorecon 3), or control points addition to wm in brainmask.mgz, then run the -cp flag. Is there any other solution left to clone the white matter (and pial voxels) left out? is brain.finalsurfs.manedit.mgz a solution? 2. Unrelated, after manual pial edits and reruning the recon-all autorecon-pial command, the pial surface generated is mostly OK, however where there have been heavy dura editing the surface looks granular. Is there a workaround (along with better hands), such as smoothing the final pia a bit more (preferably in the recon-all pipeline as to be convergent with other segmentation results). I appreciate your help, Octavian

2 4

Infinite loop in mri_convert
by Nooshin Zadeh 10 Nov '14

10 Nov '14

Dear FreeSurfer experts, I get a new message when I use "mri_convert" command to convert dicom to mgz. I used to apply this command before and there was no problem. I wonder what is the reason that I receive the following message. In addition, it does not create mgz file, and goes to infinite loop. *mri_convert "/M1/t1_se_ax_5/IM-0004-0001.dcm" "/FS/M1/mri/orig/001.mgz"* mri_convert /M1/t1_se_ax_5/IM-0004-0001.dcm /FS/M1/mri/orig/001.mgz $Id: mri_convert.c,v 1.166.2.2 2010/08/10 19:11:50 greve Exp $ reading from /M1/t1_se_ax_5/IM-0004-0001.dcm... Getting Series No Scanning Directory INFO: Found 39 files in /Users/christine/Raw_Data/M226/t1_se_ax_5 INFO: Scanning for Series Number 5INFO: found 37 files in series INFO: loading series header info. I appreciate if you can advise me. Regards, Nooshin PhD Candidate Departments of Electrical & Computer Engineering University of Miami Email: n.zadeh(a)miami.edu

3 2

mri_surfcluster and "sub-clustering"
by Sophie Maingault 10 Nov '14

10 Nov '14

Dear FS experst, I used the command mri_surfcluster. For one analysis, I obtained 6 clusters with one huge of 473 cm2 which corresponds to several "sub-clusters" because it was expanded through a large part of the cortex (covering several anatomical regions). Is there a way to separate one apparent cluster in several sub clusters by extracting several maximum vertices and not only one per cluster ? Best, Sophie Maingault – Doctorante/PhD student GIN (Groupe d'Imagerie Neurofonctionnelle) UMR 5296 – Université de Bordeaux – CNRS/CEA 146, rue Léo Saignat – Case 71 33076 Bordeaux cedex

2 1

Pial edits and wm control points.
by Octavian Lie 08 Nov '14

08 Nov '14

Dear all, I am editing recon-all segmentation results in freeview using the recon edit function. 1. Is there a way to complement the toggle between the brainmask.mgz and T1.mgz, especially at the end of pial editing, before resubmitting the saved brainmask volume to recon-all? It would be nice to be able to see the edited voxels (the final ones remaining edited, after successive deletions/clones) as a mask over the T1.mgz in freeview as the final check before sending brainmask back to segmentation. 2. As far as control points go, the FS tutorial mentions that there is a certain trial and error as to the no or density of control points to be used. Still, the example fig in the tutorial shows control points spaced about 2-5-voxels. Is this a good starting point in general? 3. Some of the segmentations need bit corrections in wm edits (filling holes), missing wm in brainmask (control points with T1.mgz as reference), and pial edits. It is OK if all of these corrections are done in a single step on brainmask.mgz with aux T1.mgz, then do recon-all -autorecon2-cp -autorecon3 -subjid? Thank you for your tips, Octavian

2 2

using T2 or flair images
by will brown 08 Nov '14

08 Nov '14

Hi there, I have completed recon-all on a subject but the output is terrible. I have since found multiple series on the CD that was supplied for the subject's scan using "dcmunpack" the output was: Found 7 unique series: 1 4 5 6 7 8 9 Subject XXXXX Date 20120114 Time 082517 Institution Hagley Radiology 1 loc 1.372 5.032 30 0.9375 ROW 244.141 /home/will/Dropbox/BrainformLtd/Freesurfer/November2014/NB/NBScan1/Z01 4 T2AXIALPROP 91.82775 4500.0 90.0 0.429688 ROW 325.39062 /<path to dicoms>/Z15 5 SAGT1FLAIR 7.64 2657.82 90 0.4688 ROW 195.312 /<path to dicoms>/Z53 6 SAGT2FRFSE 88.272 1900 90 0.4297 ROW 122.07 /<path to dicoms>/Z82 7 CORT2FLAIR 167.7 9502 90 0.4102 ROW 162.773 /<path to dicoms>/Z99 8 DWIASSET 88.5 7750 90 0.8594 COL 1953.12 /<path to dicoms>/Z141 9 Ax3DSWAN 22.876 38.2 15 0.4102 ROW 162.773 /<path to dicoms>/Z203 I'm wondering if an expert could advise me on the best next step of progress to improve the output. I only need the pial surfaces and stats. Presumably series 1 is the T1 series? Should I use T2 also? if so which one and how? Please excuse my ignorance. Thanks, Will Brown

1 0

07 Nov '14

So I am trying to use TKregister2 and I keep having this error pop up. I tried looking online for a solution. I tried adding export LIBGL_ALWAYS_INDIRECT=y to my .cshrc file but it did not help. I saw the question asked several times but never any answer to the significance of it, but I am accessing the computer remotely. Any help that could be offered would be greatly appreciated. Nicholas

3 11

Re: [Freesurfer] applying manual freesurfer edits to warped brain image
by David Grayson 06 Nov '14

06 Nov '14

Thanks Bruce, So I just need a warp field in .m3z format. Cool. As this distortion correction is proprietary, we don't actually get the warp fields, which is silly. But I was thinking since scanner distortions are subtle and coarse, it should be sufficient to nonlinearly register the two unsegmented images using a simple intensity-based volumetric registration at a coarse warp resolution. Would the mri_cvs_register tool be suitable for this if I supply the --step3 flag? Otherwise I am stuck with trying to go between Freesurfer and another set of tools. Thanks! -David Date: Sun, 2 Nov 2014 21:44:23 -0500 > From: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu <javascript:;>> > Subject: Re: [Freesurfer] applying manual freesurfer edits to warped > brain image > To: Freesurfer support list <freesurfer(a)nmr.mgh.harvard.edu <javascript:;> > > > Message-ID: <35417931-CE03-4CB8-A30E-C1F34C174B71(a)nmr.mgh.harvard.edu > <javascript:;>> > Content-Type: text/plain; charset="us-ascii" > > What format is the distortion field stored in? If you can get it into .m3z > I believe we have tools to do what you want > > > > > On Nov 2, 2014, at 9:37 PM, David Grayson <dgrayson(a)ucdavis.edu > <javascript:;>> wrote: > > > > Hi freesurfer experts, > > > > I have two T1 images from the same subject acquired at the same time > point. One of the images is distortion corrected using a calibration > phantom, put through the full freesurfer pipeline and then extensively > manually edited. The uncorrected image has had zero processing. What I > would like now is to have the freesurfer segmentations on the uncorrected > image and preserve as much as possible the manual edits. In other words, I > would like to warp the surface meshes already obtained on the distortion > corrected image to the uncorrected image. Is there already a suggested > method that is optimal for this purpose? > > > > Thanks in advance for any advice! > > > > -David > > _______________________________________________ > > Freesurfer mailing list > > Freesurfer(a)nmr.mgh.harvard.edu <javascript:;> > > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >

1 0

White matter volume
by Alshikho, Mohamad J. 06 Nov '14

06 Nov '14

Dear Bruce, My questions is just to learn and to improve my understanding. The commands recon-all and dt-recon are routinely generating the statistics at the end of there running and we can check the results in aseg.stats (the output of recon-all) and fa.stats (the output of dt-recon for FA). In those files we have the white matter volume for many different parts of the brain like the right and left cerebral white matter. If we check the volumes in those files we can see that we have differences between the numbers and I can explain that by the following: in aseg.stats the white matter volume is derived from the T1 segmentation in the structural space and in fa.stats the white matter volume is derived from the FA map in diffusion space and the difference between the volumes is the difference between the accuracy of the images in different spaces. Now: 1. If we want to study the relationship between the FA and the right hemispheric white matter volume for example, we should not care about the volumes in fa.stasts because it is not accurate?! in this case what is the optimal and correct method in freeSurfer to do such like this comparison, is it by moving the FA map to the structural space by registering the FA map to T1 or by using the volumes in aseg.stats and the metrics values from fa.stats? or we can use the fa values and the volumes from fa.stats? 2.What is the exact usage for the volumes mentioned in FA.stats 3. What is the meaning and the exact usage of min-max intensity range (intensity of what) in aseg.stats and fa.stats So Sorry for this long email but as always I am looking forward to learn from you Thanks Mohamad

2 2

Re: [Freesurfer] adding insula voxels
by Adam Mezher 06 Nov '14

06 Nov '14

Hi Ani, I actually did end up fixing the insula mis-segmentation using either control points or WM edits for many subjects (I think the one I sent you too, not sure). However, I did notice that in some subjects, the edits will push the wm line out a little too much such that minor part of the medial insula will be labeled as WM instead of insula. It is pretty minor though so don’t think it is a big deal anyway. -Adam

2 1

← Newer
1
...
9
10
11
12
13
14
15
Older →

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 November 2014