Freesurfer

freesurfer@nmr.mgh.harvard.edu

5 participants
30076 discussions

which image is the best
by Feng-Xian Yan 25 Jan '08

25 Jan '08

Hi, I have a problem about which image is the best to run. I had ran the 1.5T data in freesurfer. Because the contrast of the 1.5T data was too bad, I use the 3T data and rerun again. When I run, it says the warning, WARNING ============++++++++++++++++++++++++======================== The physical sizes are (1404.00 mm, 256.00 mm, 256.00 mm), which cannot fit in 256^3 mm^3 volume. The resulting volume will have 1404 slices. The freesurfer tools should be able to handle more than 256 slices. If you find problems, please let us know (freesurfer(a)nmr.mgh.harvard.edu) =====================++++++++++++++++++++++++=============== This problem confused me. What can I do? Could you tell me what image is freesurfer I can run and get the best results? (about dimension…. )

2 1

vertex neighbourhood
by nima＠nmr.mgh.harvard.edu 24 Jan '08

24 Jan '08

hi surfers, Imagine that I need to know what are the closest N vertices (let's say n=100) to a given vertex (i mean two dimentionsal neighourhood on surfaces) . Any idea how to do that? anybody has a written code for this? cheers nima

4 3

mri_cc segmentation fault
by Heath Pardoe 24 Jan '08

24 Jan '08

Hi, we've run into an issue running FS 4.01 on an Altix cluster (SuSE Itanium distribution - freesurfer-ia64-unknown-linux-gnu-stable4-20070918). On some subjects mri_cc falls over with a segmentation fault error (output pasted below). Running mri_cc -aseg aseg.auto.mgz -o aseg.auto_CCseg.mgz subject_id on the same subject on a different machine (Mac OS X) works ok. If we run mri_cc (on a subject that worked on the Altix) on our other machine we get identical output in recon-all.log. It seems strange that it works on some subjects and not on others. Any ideas as to what's going on? Please let me know if I need to provide any further information. Thanks Heath #-------------------------------------- #@# CC Seg Wed Jan 23 00:00:15 EST 2008 mri_cc -aseg aseg.auto.mgz -o aseg.auto_CCseg.mgz xxxx reading input aseg from aseg.auto.mgz writing aseg with cc labels to aseg.auto_CCseg.mgz reading aseg from xxxx/mri/aseg.auto.mgz reading norm from xxxx/mri/norm.mgz 43790 voxels in left wm, 72489 in right wm, xrange [117, 129] global minimum found at slice 123.0, rotations (-1.75, -2.25) final transformation (x=123.0, yr=-1.750, zr=-2.250): 0.999 0.039 -0.031 4.033; -0.039 0.999 0.001 4.772; 0.031 0.000 1.000 -3.697; 0.000 0.000 0.000 1.000; Segmentation fault (core dumped) Linux xxxx 2.6.16.53-0.16-default #1 SMP Tue Oct 2 16:57:49 UTC 2007 ia64 ia64 ia64 GNU/Linux recon-all exited with ERRORS at Wed Jan 23 00:05:46 EST 2008 -- Heath Pardoe, PhD Brain Research Institute Neurosciences Building Austin Health Banksia Street Heidelberg West, Victoria, 3081, Australia Phone: (+61 3) 9496 2576 Fax: (+61 3) 9496 4071 Email: h.pardoe(a)brain.org.au

1 0

what is the prerequisite of cc in cortparc?
by walhovd＠nmr.mgh.harvard.edu 23 Jan '08

23 Jan '08

Hi I have a dataset originally processed with the Autumn 2005 FS version (2.2?). I know need wm segs, and in order to get this, I've tried to run -cortribbon -cortparc -cortparc2 -aparc2aseg -wmparc But the cc doesn't get labeled in the cortparc, and consequently (?) not in the wmparc either. I've also tried to rerun all of -autorecon3, but w/o difference. When rerunning recon-all -all, it gets fixed, but I would like to keep as much as possible as is, since I did lots of edits due to a temporal bias field problem. I've edited the asegs in some instances, and that is not kept by default. Grateful for tips on how far down in the processing stream I need to go to get correct wmparcs w/cc. Thanks, Kristine

4 6

comparison of twins
by Thomas Doering 23 Jan '08

23 Jan '08

Hi, i am wondering how i can do a comparison of just two subjects (twins), calculating differences between them in cortical thickness, gm-volume .etc... I tried to find some paper with a processing pipline for this but without success... Perhaps somebody has a hint for me? Thanks, Thomas

2 1

losing aseg edits
by Michael Harms 23 Jan '08

23 Jan '08

Hello, Kristine's earlier post got me wondering why she was "losing" any previous aseg edits if her data were re-processed under FS v4, since it was my simple understanding that all edits (e.g., control points, brainmaks, talairach, white matter, aseg) are intended to be preserved, unless you explicitly set one of the "clean" flags. Looking at the recon-all script for FS v4.0.2, I now see that any aseg edits will indeed be lost if the corpus callosum segmentation has not been previously already run (e.g., say if you are re-processing data originally processed under FS v3). Namely, if the file "aseg.auto_CCseg.mgz" does not exist, then variable 'AsegDiff' gets set to 0, which in turn results in aseg.auto_CCseg.mgz being subsequently copied to aseg.mgz, thus losing any previous edits to the aseg.mgz. Is this the intended behavior? Is there an easy way to get the CC segmentation while still preserving any previous aseg edits? cheers, Mike H. -- Michael Harms, Ph.D. -------------------------------------------------------------------- Conte Center for the Neuroscience of Mental Disorders Washington University School of Medicine Department of Psychiatry, Box 8134 Renard Hospital, Room 6615 Tel: 314-747-6173 660 South Euclid Ave. Fax: 314-747-2182 St. Louis, MO 63110 Email: mharms(a)wustl.edu --------------------------------------------------------------------

2 1

?h.volume files
by bell0368＠umn.edu 23 Jan '08

23 Jan '08

Freesurfers, The wiki mentions one of the "known issues" with the current release is that recon-all does not create ?h.volume files. I have been using the following command to create what appear to be volume files for cortical and wm parcellations. Is the problem just that these files are not created automatically? mri_extract_label aparc+aseg.mgz (number corresponding to the desired label) (name of the output label/volume file) Chris Bell

2 1

Re: [Freesurfer] recommended scan parameters
by Sebastian Moeller 22 Jan '08

22 Jan '08

Hi Michael, On Jan 22, 2008, at 4:18 PM, Michael Chee wrote: > Hi Sebastian and Bruce, > > I'll respond by the end of the week as I have some things to tend to > presently. Cool, thanks a lot and no hurry (no(t much) time this week either for testing/acquiring new anatomical scans myself). ahoi & thanks Sebastian > > Mike > > >> From: Sebastian Moeller <sebastian.moeller(a)lur.rwth-aachen.de> >> Date: Tue, 22 Jan 2008 15:10:15 +0100 >> To: Bruce Fischl <fischl(a)nmr.mgh.harvard.edu> >> Cc: Mailing Freesurfer List <freesurfer(a)nmr.mgh.harvard.edu>, >> Andre van der >> Kouwe <andre(a)nmr.mgh.harvard.edu>, Mike Chee <mchee(a)pacific.net.sg> >> Subject: Re: [Freesurfer] recommended scan parameters >> >> Hi Bruce, >> >> thanks for your quick response. >> >> On Jan 22, 2008, at 3:06 PM, Bruce Fischl wrote: >> >>> Hi Sebastian, >>> >>> we put together some Allegra-specific postprocessing tools that >>> helped for Mike Chee (ccd). Maybe Mike can give you the benefit of >>> his experience? >> That would be marvellous. Thanks in advance. >> >> ahoi >> Sebastian >> >>> >>> cheers, >>> Bruce >>> >>> On Tue, 22 Jan 2008, Sebastian Moeller wrote: >>> >>>> Dear Experts, >>>> >>>> after ages of monkey scanning, we recently also included human >>>> subjects in our study. I would like to use freesurfer for surface >>>> reconstruction, mainly as "canvas" for the display of activation >>>> maps. >>>> While functional scans seem to be reasonable, the anatomicals >>>> taken on our Siemens Allegra (3T) with the default Siemens head >>>> coil are not well reconstructed by freesurfer (pial surface is >>>> both under- and over-inclusive, white matter in temporal pole is >>>> mis-classified as gray). Just in case it matters, test >>>> reconstructions were performed using the 64 bit centos_4 >>>> FreeSurfer 4.0.2 build on SuSUE 10.1 64 bit (and no errors in from >>>> the recon-all -autorecon1 and -autorecon2 stages). I assume that >>>> this is a driver's error on my part, as switching to the sequence >>>> parameters given on http://www.nmr.mgh.harvard.edu/~andre/ >>>> Sequences/SagMPR8Min3T.html improves the quality of the >>>> reconstruction considerably. (Even though with our sequence we can >>>> actually set the FFT scaling factor.) >>>> Now I would be happy to know whether these protocol parametrs are >>>> still the recommended set for anatomical scans on an Allegra >>>> (using only standard "off-site" Siemens sequences)? I would also >>>> be grateful for anybodies input on the matter of "optimising" the >>>> raw data for the surface reconstruction, so manual interventions >>>> can be minimised. >>>> >>>> >> >> -- >> Sebastian Moeller >> >> Tel.: 04 21 - 2 18 - 78 38 oder 96 91 >> Fax.: 04 21 - 2 18 - 90 04 >> GSM: 0 15 77 - 1 90 31 41 >> moeller(a)brain.uni-bremen.de >> >> AG Kreiter / FB 2 >> Institut fuer Hirnforschung III >> Abteilung Theoretische Neurobiologie >> Universitaet Bremen >> Biogarten >> Hochschulring 16a >> Postfach 33 04 40 >> 28359 Bremen >> >> > > -- Sebastian Moeller Tel.: 04 21 - 2 18 - 78 38 oder 96 91 Fax.: 04 21 - 2 18 - 90 04 GSM: 0 15 77 - 1 90 31 41 moeller(a)brain.uni-bremen.de AG Kreiter / FB 2 Institut fuer Hirnforschung III Abteilung Theoretische Neurobiologie Universitaet Bremen Biogarten Hochschulring 16a Postfach 33 04 40 28359 Bremen

1 0

recommended scan parameters
by Sebastian Moeller 22 Jan '08

22 Jan '08

Dear Experts, after ages of monkey scanning, we recently also included human subjects in our study. I would like to use freesurfer for surface reconstruction, mainly as "canvas" for the display of activation maps. While functional scans seem to be reasonable, the anatomicals taken on our Siemens Allegra (3T) with the default Siemens head coil are not well reconstructed by freesurfer (pial surface is both under- and over-inclusive, white matter in temporal pole is mis-classified as gray). Just in case it matters, test reconstructions were performed using the 64 bit centos_4 FreeSurfer 4.0.2 build on SuSUE 10.1 64 bit (and no errors in from the recon-all -autorecon1 and -autorecon2 stages). I assume that this is a driver's error on my part, as switching to the sequence parameters given on http:// www.nmr.mgh.harvard.edu/~andre/Sequences/SagMPR8Min3T.html improves the quality of the reconstruction considerably. (Even though with our sequence we can actually set the FFT scaling factor.) Now I would be happy to know whether these protocol parametrs are still the recommended set for anatomical scans on an Allegra (using only standard "off-site" Siemens sequences)? I would also be grateful for anybodies input on the matter of "optimising" the raw data for the surface reconstruction, so manual interventions can be minimised. -- Sebastian Moeller Tel.: 04 21 - 2 18 - 78 38 oder 96 91 Fax.: 04 21 - 2 18 - 90 04 GSM: 0 15 77 - 1 90 31 41 moeller(a)brain.uni-bremen.de AG Kreiter / FB 2 Institut fuer Hirnforschung III Abteilung Theoretische Neurobiologie Universitaet Bremen Biogarten Hochschulring 16a Postfach 33 04 40 28359 Bremen

2 2

Re: [Freesurfer] Phillips 3T
by Roberto Roiz Ssantiáñez 21 Jan '08

21 Jan '08

Hi When I bring the images up with tkmedit everything is messed up. When I open the coronal view it brings up the axial view, when I open the axial view it brings up the sagital view and when I open the sagital view it brings up the coronal view. As I mentioned before this happens when I acquired the sequence in the sagital plane. When the sequence is acquired in the axial plane everything runs ok Regards Roberto Roiz > if you bring them up in tkmedit does it show them properly? That is, when > it says "coronal" etc... is it really? > > And please post these to the list so other people can answer :) > > Bruce > On Sat, 19 > Jan 2008, Roberto Roiz > Ssantiáñez wrote: > >> Hi Bruce, >> I'm using >> freesurfer-Linux-rh9-stable-pub-v4.0.2-full.tar.gz run on Ubuntu. >> The error I get when runnig a sequence acquired on the sagital plane is >> is: >> >> Sat Jan 19 11:38:02 CET 2008 >> talairach_avi done >> #-------------------------------------------- >> #@# Talairach Failure Detection Sat Jan 19 11:38:02 CET 2008 >> /usr/local/freesurfer/subjects/pilot2_sag/mri >> >> talairach_afd -T 0.005 -xfm transforms/talairach.xfm >> >> ERROR: talairach_afd: Talairach Transform: transforms/talairach.xfm >> ***FAILED*** (p=0.0000, pval=0.0000 < threshold=0.0050) >> Linux roberto-desktop 2.6.22-14-generic #1 SMP Tue Dec 18 08:02:57 UTC >> 2007 i686 GNU/Linux >> >> recon-all exited with ERRORS at Sat Jan 19 11:38:02 CET 2008 >> >> >> >> However when the same sequence is acquired on the axial plane the >> recon-all runs perfectly. >> I have converted both sequences using MRICRO. I attached an image of >> both >> the sagital and the axial sequences. >> >> Regards >> Roberto Roiz >> >> >> >>> Hi Roberto, >>> >>> what version of FS are you using? Usually this isn't a problem for us. >>> You can drop an example on our filedrop if you want and we'll take a >>> look. >>> >>> cheers, >>> Bruce >>> On Wed, 2 Jan 2008, Ron KILLIANY wrote: >>> >>>> Dear Roberto, >>>> >>>> I'm glad to hear that the sequence ran well. I'm going to include >>>> Bruce >>>> Fischl on this e-mail to see if he can help you with the orientation >>>> problem you seem to be having. To my knowledge we haven't experienced >>>> that problem. >>>> >>>> I hope you had a happy new year celebration! >>>> >>>> --ron >>>> >>> >>

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