Freesurfer

freesurfer@nmr.mgh.harvard.edu

16 participants
30084 discussions

longitudinal processing
by Nicola Hobbs 16 Oct '08

16 Oct '08

Hi there, We would like to do some longitudinal processing. Our baseline scans have been run through v4.0.3, and we spent a long time manually editing the cortical segmentations. Ideally we would like to use these segmentations as the reference for our longitudinal processing (v4.1.0), but we are worried that using different versions for the same study could introduce errors. Could you advise us on whether we need to re-run our baseline data through v4.1.0 before we attempt any longitudinal processing? Many thanks for your help, Nicola -- Nicola Hobbs Dementia Research Centre National Hospital for Neurology and Neurosurgery Queen Square London WC1N 3BG +44 (0)845 155 5000 ext. 723839

2 1

structural ROI analysis
by Wiegand, Katrin 16 Oct '08

16 Oct '08

Hi, I'm using freesurfer v4.0.5 and I am trying to do a ROI analysis using func2roi-sess. However, there seems to be some problem with the structural label (created with tkmedit). I.e., when running func2roi-sess only with a structural constraint, the report includes following messages INFO: vol2maskavg: nhits = 0 WARNING: there were no voxels in the input mask > 0.5 (checking the final mask of the output also reveals no voxels being included). Furthermore, when I'm trying to use -raw (instead of averaged data defined by -analysis) I get the following error message: ERROR: no analysis specified Any suggestions how to solve the problems? Thanks, Katrin

1 0

Re: recon-all with voxel sizes other than 1mm^3
by Anderson Winkler 15 Oct '08

15 Oct '08

Hi, I edited the recon-all, omitting the --conform option, and the pipeline failed shortly after. So, is there a way to run the volume processing pipeline (in particular, the subcortical segmentation), without resampling to 1mm^3 voxels, keeping the native format 0.8^3 mm^3? Anderson

2 1

new Freesurfer release v4.1.0
by Nick Schmansky 14 Oct '08

14 Oct '08

Freesurfer users, There is a new release of freesurfer available, v4.1.0. This is a bug- fix release, described on this wiki page: https://surfer.nmr.mgh.harvard.edu/fswiki/ReleaseNotes The download page is here: https://surfer.nmr.mgh.harvard.edu/fswiki/Download Mainly this release is of interest to those waiting for the updated longitudinal analysis stream. Nick

1 0

qdec -- Cannot allocate memory
by Richard Frye 14 Oct '08

14 Oct '08

I am running qdec on a Mac Pro OS X 10.5.5 with dual quad pentiums mirrored RAID and 16GB memory. I have followed the instruction for creating the qdec table, loaded the data in successfully and requested an analysis with two continuous variance. The following is the final output before it crashes because it cannot allocate memory. I have the maximum memory in my machine. What can I do? $Id: mri_glmfit.c,v 1.138.2.1 2007/09/12 15:38:19 nicks Exp $ cwd /Volumes/share/MRI_Data cmdline mri_glmfit --y /Volumes/share/MRI_Data/qdec/Untitled/y.mgh --fsgd /Volumes/share/MRI_Data/qdec/Untitled/qdec.fsgd --glmdir /Volumes/share/MRI_Data/qdec/Untitled --surf fsaverage lh --label /Volumes/share/MRI_Data/fsaverage/label/lh.cortex.label --C /Volumes/share/MRI_Data/qdec/Untitled/contrasts/Avg-Intercept-thickness.mat --C /Volumes/share/MRI_Data/qdec/Untitled/contrasts/Avg-thickness-RL-Cor.mat --C /Volumes/share/MRI_Data/qdec/Untitled/contrasts/Avg-thickness-RC-Cor.mat sysname Darwin hostname Macintosh-2.local machine i386 user stevenhecht FixVertexAreaFlag = 1 UseMaskWithSmoothing 1 OneSampleGroupMean 0 y /Volumes/share/MRI_Data/qdec/Untitled/y.mgh logyflag 0 usedti 0 FSGD /Volumes/share/MRI_Data/qdec/Untitled/qdec.fsgd labelmask /Volumes/share/MRI_Data/fsaverage/label/lh.cortex.label maskinv 0 glmdir /Volumes/share/MRI_Data/qdec/Untitled DoFFx 0 Creating output directory /Volumes/share/MRI_Data/qdec/Untitled Loading y from /Volumes/share/MRI_Data/qdec/Untitled/y.mgh Matrix condition is 131.103 Found 149955 points in label. Pruning voxels by frame. Found 149955 voxels in mask Saving mask to /Volumes/share/MRI_Data/qdec/Untitled/mask.mgh search space = 59659.8 DOF = 32 Starting fit and test 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Fit completed in 0.0612333 minutes Computing spatial AR1 on surface Residual: ar1mn=0.997008, ar1std=0.001276, gstd=6.640355, fwhm=15.636842 Writing results Avg-Intercept-thickness maxvox sig=23.8959 F=850.546 at index 105774 0 0 seed=1224021971 Avg-thickness-RL-Cor maxvox sig=3.8835 F=18.8999 at index 129111 0 0 seed=1224021971 Avg-thickness-RC-Cor maxvox sig=-3.9267 F=19.2034 at index 152642 0 0 seed=1224021971 mri_glmfit done ninputs = 3 Checking inputs nframestot = 3 Allocing output Writing to /Volumes/share/MRI_Data/qdec/Untitled/contrasts.sig.mgh Reading in average area /Volumes/share/MRI_Data/fsaverage/surf/lh.white.avg.area.mgh Reading in average area /Volumes/share/MRI_Data/fsaverage/surf/lh.pial.avg.area.mgh qdec.bin(9066,0xa049cfa0) malloc: *** mmap(size=1953464320) failed (error code=12) *** error: can't allocate region *** set a breakpoint in malloc_error_break to debug TAGskip: failed to calloc 1953461248 bytes! Cannot allocate memory Best Wishes, Richard E. Frye, M.D., Ph.D., F.A.A.P. Assistant Professor of Pediatrics and Neurology University of Texas Medical School at Houston 7000 Fannin -- UCT 2478 Houston, TX 77030 Office: 713-500-3245 e-mail: richard.frye(a)gmail.com

2 1

recon-all : in a raw or parallel?
by asaf achiron 14 Oct '08

14 Oct '08

dear all, i would like to run freesurfer during the weekend for 2 patients. what is better, to run recon-all after recon-all or to run parallel on second terminal? thank you, Asaf

3 2

help - Intensity Normalization2 error
by Andreas Engvig 14 Oct '08

14 Oct '08

Hi, We have previously run recon-all -all on 72 subjects using Freesurfer v. 4.01. After routine edits we reprocessed the subjects. I now want to run the scans again using the updated FS v. 4.05 using the recon-all -all command. We want to have them processed in the latest version in order to do longitudinal analyses. However, when I run the v. 4.01-processed scans again, using v.4.05 I get the following error in the normalization2 part of the autorecon2 step: #@# Intensity Normalization2 Fri Oct 10 22:10:06 CEST 2008 /work/1531087.undefined.d/2003A/mri mri_normalize -aseg aseg.mgz -mask brainmask.mgz norm.mgz brain.mgz using segmentation for initial intensity normalization reading from norm.mgz... normalizing image... removing outliers in the aseg WM... 68543 control points removed building Voronoi diagram... mriMarkUnmarkedNeighbors: all inputs must be MRI_UCHAR using MR volume brainmask.mgz to mask input volume... using wm (107) threshold 97.0 for removing exterior voxels Linux compute-3-3.local 2.6.9-67.0.15.ELsmp #1 SMP Thu May 8 10:50:20 EDT 2008 x86_64 x86_64 x86_64 GNU/Linux recon-all exited with ERRORS at Fri Oct 10 22:10:51 CEST 2008 -- The same error occurs for all subjects. The error only occurs when I run the subjects on our computing cluster. I have tried to run a couple of scans on a local machine, using the same version. Running locally, I don't get the same error. We use centOS4 64bit. Any suggestions on how to solve this? Thanks ! Best regards, Andreas Engvig -- Andreas Engvig Center for the Study of Human Cognition University of Oslo PB 1094 Blindern 0317 Oslo Cell: +47 988 58 770

1 0

Question about mri_ca_normalize
by John Sheppard 13 Oct '08

13 Oct '08

Hey Freesurfers, Couple question about 2 FS commands: first is mri_normalize, a command used in -autorecon2 pipeline. 1) Can someone tell me exactly what flags to pass this command in order for it to be as aggressive as possible in bumping up intensities, in an attempt to end up with more expanded grey/white and pial surfaces? (I'm currently losing lateral cortex in surfaces, even with extensive manual control points.) Here's a link to the command info: http://surfer.nmr.mgh.harvard.edu/fswiki/mri_normalize Next question is about mri_ca_normalize, another command used in -autorecon2: 2) From what I gather, the way to ensure that mri_ca_normalize is being as aggressive as possible in declaring white matter is by setting the flag -p to 1.0 - Is that right? Are there any other options that could help in this regard? Link: http://surfer.nmr.mgh.harvard.edu/fswiki/mri_ca_normalize My issue is sagittally-acquired data with bad inhomogeneity - namely cortical grey matter is far lateral cortex that is not as intense as it needs to be. Normalization needs to be bumping up the intensity in these regions and instead it's still lowering the intensity, resulting in loss of cortex in pial surface. Adding control points to ensure white/grey boundary is correct tends to expand the pial surface when segmentation doesn't get enough WM, which helps me, but it's not enough. Thus I want to ensure I am maximizing the potential for WM (and indirectly, pial) surface size through these normalization commands. I'm also looking at the mri_nu_correct.mni command which was taken from MNI - haven't had much luck there but I've been messing with the iteration number (and trying removing it completely). Thanks in advance! -- John Sheppard Undergraduate Student (Class of 2010) Department of Biomedical Engineering Northwestern University Phone: (630) 957-8321 E-mail: john-sheppard(a)northwestern.edu

2 1

recon-all with voxel sizes other than 1mm^3
by Anderson Winkler 13 Oct '08

13 Oct '08

Dear all, Is there a way to have all the results from recon-all in the native image space (in our case, 0.8x0.8x0.8mm), rather than conformed to 1mm^3? Looking at the script, I'm worried that replacing the --conform to --noconform option to mri_convert may prevent the downstream steps to work correctly. Thanks in advance, Anderson PS: I saw the recent message from João to the list and the follow up, but I think that this is a different question anyhow. PS 2: Sorry for sending twice, but I think that the previous doesn't open properly in the Pipermail page. :-)

1 0

recon-all with voxel sizes other than 1mm^3
by relkniw 13 Oct '08

13 Oct '08

1 0

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