Freesurfer August 2005

freesurfer@nmr.mgh.harvard.edu

42 participants
72 discussions

Re: [Freesurfer] mgz conversion
by Rudolph Pienaar 18 Aug '05

18 Aug '05

I should have mentioned as a double check that you can type the following: rpm -qa | grep gsl to see if any 'rpm' (the package management system) files installed on your system match the gsl. Probably you won't find anything. The easiest way to install software for a RedHat system is to use 'rpm' files. I have added 'rpm' files for the gsl libraries to the Martinos Center's "File Drop" application. This app will send you an automatic email with instructions how to download these files (you should have received the notification by now). Follow the instructions you've received and download the files (there are two - gsl-devel-1.5-1.i386.rpm and gsl-1.5-1.i386.rpm) to your local box. OK - for the following, I'm assuming that you have root access to your local machine. Change to the root user: su - (and enter the root password). Change to the directory where you downloaded the gsl rpm files. Say you downloaded them to /tmp: cd /tmp Now install them with rpm -Uhv *.rpm and you'll see a short report as each rpm is installed. You should be all set now. Good luck, -=R -- Rudolph Pienaar, M.Eng, D.Eng / email: rudolph(a)nmr.mgh.harvard.edu MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging 149 (2301) 13th Street, Charlestown, MA 02129 USA

1 0

using Freesurfer with external storage media
by Glenn Lawyer 17 Aug '05

17 Aug '05

Hi, I remeber a discussion here a month or so back about using Freesurfer with USB hard drives. People were having problems because the hard drive changed file names to lower case (i.e. cor instead of COR). The solution posted to the list was a script that would rename all cor files to uppercase. There is an easier way. You can reformat the drive with an ext3 file system. Ext3 is the (default) Linux file system, and provides support for case-sensitive file names (as well as permissions, links, etc). So once re-formatted you will no longer have the re-naming problem. Two warnings. 1) Reformatting will destroy ALL data on the drive; be prepared before you do this. 2) Windows cannot read ext3 file systems. Reformatting will make your drive unreadable to any Windows-based machine. To do it, as root: umount /media/usbharddisk # make sure the drive is not mounted mke2fs -j [device /dev] # reformat drive where [device /dev] is the link in the /dev directory where the drive is located. It should be something like /dev/sda1 or /dev/sdb1. You can tell by looking at the output of dmesg. Look for a section like this: -------------------- begin terminal output --------------------------- user@machine $ dmesg .......... usb 1-4: new high speed USB device using ehci_hcd and address 3 Initializing USB Mass Storage driver... scsi2 : SCSI emulation for USB Mass Storage devices usb-storage: device found at 3 usb-storage: waiting for device to settle before scanning usbcore: registered new driver usb-storage USB Mass Storage support registered. Vendor: Maxtor Model: OneTouch Rev: 0201 Type: Direct-Access ANSI SCSI revision: 00 SCSI device sdb: 490232832 512-byte hdwr sectors (250999 MB) sdb: assuming drive cache: write through SCSI device sdb: 490232832 512-byte hdwr sectors (250999 MB) sdb: assuming drive cache: write through sdb: sdb1 <---- LOOK HERE Attached scsi disk sdb at scsi2, channel 0, id 0, lun 0 usb-storage: device scan complete ----------------------------end terminal output------------------ The key line is sdb: sdb1 , which tells me that my disk is linked via sdb1, so to reformat I type mke2fs -j /dev/sdb1 +glenn -- I want to go farther, farther! A. E. L. > Glenn Lawyer < > +352 061 967 244 < > Instituttgruppe for psykiatri < > Postboks 1130 Blindern < > 0318 Oslo < < http://folk.uio.no/davidgl > <><><><><><><><><><><><><><><><><:)

2 1

.tcl file for tksurfer
by Katy Thakkar 17 Aug '05

17 Aug '05

Hi, I'm trying to write a little .tcl script to automate loading labels in tksurfer. Below is the script--I'm using the line 'setfile label /space/ventzl/1/users/SUBJECTS_DIR/mano002/label/dACC-rh.label', but the label is not loading. Does anyone why this might be? Thanks! Katy set val /autofs/space/balagula_002/users/EMexec/fMRI/mano002/bold/EMexec//PSvfix/minsig-0-rh.w sclv_read_binary_values 0 set curv rh.curv read_binary_curv; set curvflag 1; source /usr/local/freesurfer/beta/lib/tcl/readenv.tcl open_window sclv_set_current_field 0 setfile label /space/ventzl/1/users/SUBJECTS_DIR/mano002/label/dACC-rh.label redraw

2 1

RE: [Freesurfer] Editing labels
by Fornito, Alexander 16 Aug '05

16 Aug '05

Hi Kevin, Apologies for the delay. The output is below.Is it not assiging enough vertices to the label after it has been saved in tkmedit? tksurfer 1000027 lh smoothwm surfer: current subjects dir: /data/flanders/work/alex/freesurfer/subjects surfer: not in "scripts" dir ==> using cwd for session root surfer: session root data dir ($session) set to: surfer: /data/flanders/work/alex/freesurfer/subjects surfer: /data/flanders/work/alex/freesurfer/subjects/1000027/mri/T1/COR-.info INFO: Src volume /tmp/mritotal_17663/orig_8_dxyz.mnc INFO: Target volume /usr/local/mni/bin/../share/mni_autoreg/average_305_8_dxyz.mnc INFO: Volume /tmp/mritotal_17663/orig_8_dxyz.mnc cannot be found. DimSizes: 3, 86, 110, 68, 1 DataType: 1 MINC Linear Transform 2.0000 0.0000 0.0000 -86.0950 0.0000 2.0000 0.0000 -126.5100 0.0000 0.0000 2.0000 -58.2500 0.0000 0.0000 0.0000 1.0000 Done reading minc surfer: vertices=148018, faces=296032 surfer: single buffered window surfer: using interface /usr/local/freesurfer/lib/tcl/tksurfer.tcl Reading /usr/local/freesurfer/lib/tcl/tkm_common.tcl Reading /usr/local/freesurfer/lib/tcl/tkm_wrappers.tcl Reading /usr/local/freesurfer/lib/tcl/fsgdfPlot.tcl Using /usr/local/freesurfer/lib/Linux/libtclfsgdf.so Reading /usr/local/freesurfer/lib/tcl/tkUtils.tcl Read /usr/local/freesurfer/surface_labels.txt, found 95 structures Successfully parsed tksurfer.tcl % 1199 unassigned vertices in label - building spatial LUT... assigning vertex numbers to label... -----Original Message----- From: Kevin Teich [mailto:kteich@nmr.mgh.harvard.edu] Sent: Fri 8/12/2005 7:56 AM To: Fornito, Alexander Cc: freesurfer(a)nmr.mgh.harvard.edu Subject: RE: [Freesurfer] Editing labels The internal representation of the label format is treated differently between tkmedit and tksurfer. Labels are lists of RAS coords. When you load one into tkmedit, it converts each RAS coord to a voxel index for its internal data structure. When you save it out, it converts the voxel index to an RAS coord, but this can be different than the original RAS coord because it is always the corner of the voxel, where as the original point can be anywhere inside the voxel. However, when tksurfer loads the label, even though the RAS coords in the label don't match exactly with the RAS coords of the surface vertices, it should find the closest vertex coord for each RAS coord, and add that to the label. Can you send the output that tksurfer makes to the shell when you load a label that becomes fragmented like this? On Thu, 11 Aug 2005, Fornito, Alexander wrote: > Hi Kevin, > It shouldn't. > I created the labels by the doing the following: > 1 - manully raced an ROI using another program (Analyze) in volume space and saved the ROI as a greyscale mask in analyze format > 2 - used mri_convert to convert both the original T1 and ROI to COR format. > 3 - Generated the surface of the T1 and loaded the ROI as an overlay in tksurfer > 4 - Filled the overlay using custom fill and saved it as a label. > > I'm not sure how I would check the image dimensions of the label, but I assume step 2 would have put it into 1mm3, which would then convert to 1mm2 when projected onto the surface (?). > > > -----Original Message----- > From: Kevin Teich [mailto:kteich@nmr.mgh.harvard.edu] > Sent: Thu 8/11/2005 5:18 AM > To: Fornito, Alexander > Cc: freesurfer(a)nmr.mgh.harvard.edu > Subject: Re: [Freesurfer] Editing labels >> I've been trying to edit some labels in tkmedit and have noticed that if >> I save a label in tkmedit and then project it back onto the surface in >> tksurfer, it appears fragmented (ie., speckled). This happens even if I >> save a label that was created in tksurfer in tkmedit without even >> editing it (ie., create label in tksurfer, load it in tkmedit, save it >> in tkmedit (without editing), and then load it back in tksurfer). I >> don't understand why this should happen, particulalry in the latter >> case, as the label should remain unchanged. Can anyone explain why this >> might happen? Thanks, Alex > > Does your data have a voxel resolution different than 1mm squared? > > > -- Kevin Teich

1 0

Label colours
by Fornito, Alexander 15 Aug '05

15 Aug '05

Hi, Can labels be loaded in different colours in tkmedit? Alexander Fornito M.psych (Clin. Neuro.)/PhD candidate Melbourne Neuropsychiatry Centre and Department of Psychology The University of Melbourne alexander.fornito(a)wh.org.au

2 1

mri_convert
by Colleen Kristen Tuffy 15 Aug '05

15 Aug '05

Hello everyone, I'm having trouble running mri_convert on the dicom images coming off our Siemens Allegra 3T head-only scanner. With this command line: > mri_convert 5308.dcm /Volumes/Storage/data/freesurfer/subjects/scfr/mri/orig/001.mgz I get this error: > unknown file type for file 5308.dcm Setting the input type to Siemens: > mri_convert -it siemens 5308.dcm /Volumes/Storage/data/freesurfer/subjects/scfr/mri/orig/001.mgz I get this error: > reading from 5308.dcm... > ERROR: could not parse /Volumes/Storage/data/freesurfer/subjects/scfr/mri/orig/01/5308.dcm > ERROR: siemensRead(): /Volumes/Storage/data/freesurfer/subjects/scfr/mri/orig/01/5308.dcm And trying the siemens_dicom format (that perhaps I should have started with, in the first place): > mri_convert -it siemens_dicom 5308.dcm /Volumes/Storage/data/freesurfer/subjects/scfr/mri/orig/001.mgz I get: > reading from 5308.dcm... > ERROR: /Volumes/Storage/data/freesurfer/subjects/scfr/mri/orig/01/5308.dcm is not a siemens DICOM file > INFO: loading series header info. Followed by a series of malloc errors similar to this one: > mri_convert(5315) malloc: *** Deallocation of a pointer not malloced: 0x3fc7; This could be a double free(), or free() called with the middle of an allocated block; Try setting environment variable MallocHelp to see tools to help debug followed by a Bus error. What am I doing wrong? We are now running Tiger (10.4.2) on our G5. I stuck a copy of the first image in the series here: http://freylab.uoregon.edu/freesurfer/5308.dcm Please let me know if this input format is not supported, yet, or what additional information I can provide to help solve this problem. I can probably come up with a work-around, like going through NIfTI format, if necessary, but it would be neat to be able to skip the additional step and import the .dcm files directly into FreeSurfer. Thanks for your help or advice, Colleen Tuffy Human Neuroimaging Lab University of Oregon

1 0

getting data from specific points
by Goldman, Aaron (NIH/NIMH) 15 Aug '05

15 Aug '05

Hi. In order to check for outliers, I'd like to extract thickness values for all subjects at a couple chosen points (probably into a text file, so I could import it into SPSS). What would be the best way to do this? Thanks, -Aaron- Aaron L. Goldman Post-Bacc IRTA Genes, Cognition and Psychosis Program, NIMH Building 10, Room 3C108 Phone: (301) 435-0944 Email: aarong(a)mail.nih.gov <mailto:aarong@mail.nih.gov>

1 0

statistical analysys
by Bagnato, Francesca (NIH/NINDS) 15 Aug '05

15 Aug '05

Dear Free-surfer Users: when comparing cortical thickness measurements between patients and healthy people by the means of independent -tests, would you recommend correcting by Bonferroni method and consider as significant a p-value given by 0.05/82 or going ahead and just forget about the correction? Any suggestion on that would be greatly appreciated. Thanks! - francesca Francesca Bagnato, MD NIB-NINDS-NIH 10 Center Drive Building 10, Room 5B16 Bethesda, MD, 20892 USA ph #: 001-301-402.6391 fax #: 001-301-402.0373

2 1

recon-all
by antao＠itsa.ucsf.edu 15 Aug '05

15 Aug '05

Hi, After processing the data in recon-all stage2, I don't want to repeat stage2 in recon-all stage3. Is there any way to do that? Thx, Antao

2 1

mgz conversion
by David Soscia 15 Aug '05

15 Aug '05

Hey Guys, I have been converting dicom files to COR for the past month and all has been working fine. Now I am trying to convert dicom files to mgz format and I am getting an error every time. I type in the same thing I did for the COR conversion, except I replace "-ot cor" with "-ot mgz." The error is get is "mri_convert: unknown output volume type mgz" (or seomthing along those lines). I think my computer isn't recognizing the mgz format because I've never used it for anything before. Anyone know how I can fix this so it works properly? thanks, Dave David A. Soscia Neurosciences Advanced Imaging Research Center 47 New Scotland Ave. Albany, NY 12208-3479 Building S Rm. 132 P- (518)-262-0826 C- (518)-469-8317 sosciad(a)mail.amc.edu

3 2

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Freesurfer August 2005