Freesurfer June 2014

freesurfer@nmr.mgh.harvard.edu

155 participants
234 discussions

Re: [Freesurfer] FSFAST
by Douglas N Greve 04 Jun '14

04 Jun '14

Hi Walid, please remember to send email to the list and not to us personally. Thanks! It is possible to use motion as a nuisance regressor. I don't like to do so because you sometimes get some motion that is correlated with the task. This can greatly change the results. Adding the MC nuisance variable often does not help much with the residual motion signal. So it tends to hurt and not help much so I don't recommend it. doug On 06/04/2014 06:25 AM, Walid Yassin wrote: > Dear Douglas, > > I have a question about FSFAST, is it possible to tell me why is it not ok to use motion as nuisance regressor? > > Thank you in advance, > > Best, > Walid > > -- Douglas N. Greve, Ph.D. MGH-NMR Center greve(a)nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422 Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/

1 0

Orthogonalising covariates
by Saloni Sharma 04 Jun '14

04 Jun '14

Dear Freesurfers, does qdec allows you to orthogonalise one covariate relative to the other? Regards, Saloni

2 1

Re: [Freesurfer] v.5.3 Longitudinal Analysis on v.5.0 processed scans
by Martin Reuter 04 Jun '14

04 Jun '14

Hi Elijah, Just create a link ln -s pathtofsaverage ./ Usually this is done during recon-all Sent via my smartphone, please excuse brevity. -------- Original message -------- From: Elijah Mak <fkm24(a)medschl.cam.ac.uk> Date:06/04/2014 7:17 AM (GMT-05:00) To: Martin Reuter <mreuter(a)nmr.mgh.harvard.edu>,freesurfer(a)nmr.mgh.harvard.edu Subject: Re: [Freesurfer] v.5.3 Longitudinal Analysis on v.5.0 processed scans Hi Martin, We corresponded about the longitudinal pipeline earlier. Your help has been much appreciated. I am now at the stage where I am ready to perform QDEC analyses using the 2-stage method. In my Subject DIR, I have all the longs (Tp1 and 2) and their corresponding bases. However, when i run the following command long_mris_slopes --qdec ./qdec/long.qdec.table.dat --meas thickness --hemi lh --do-avg --do-rate --do-pc1 --do-spc --do-stack --do-label --time years --qcache fsaverage --sd $SUBJECTS_DIR It says that fsaverage is missing. Can I simply copy the fsverage folder from the default Freesurfer "subjects" folder? Many thanks! Best Regards, Elijah On Wed, Jan 15, 2014 at 10:15 PM, Martin Reuter <mreuter(a)nmr.mgh.harvard.edu> wrote: > Hi Elijah, > > > On 01/15/2014 04:10 PM, Elijah Mak wrote: > > Hi Martin, > > Thank you for your help. > > By re-processing, do you mean running recon-alls all over again? > > > Yes. 5.0 is old and it is better to run everything with 5.3 from beginning > to end. If you have a lot of data or a lot of edits, you can try running > -base and -long in 5.3 and keep the independent (cross sectional) stuff in > 5.0, but I don't recommend mixing versions. > > > > Could you elaborate on the advantage of creating a new subjectsdir and > copying over the orig dirs for each subject? > > This way you keep a backup copy of your old 5.0 processed data (to lookup > edits etc). If you have a script that calls recon-all with the -i flag for > inputs, you don't need to copy the mri/orig dirs (it will grab the data from > the dicoms again). Copying the orig dirs and dropping -i will skip that > step. > > > On a related note, would it be better to run the full recon-all pipeline for > each subject? I have been running batches of subjects at each stage (i.e. > autorecon1, autorecon2, and finally autorecon3). > > Yes, full pipeline (-all) on all subjects is what I usually do. You can run > individual steps and check intermediate results if you like, but I usually > run everything and check everything at the end. > > Best, Martin > > > > Thank you again. > > Best Wishes, > Elijah > > > > > On Wed, Jan 15, 2014 at 9:02 PM, Martin Reuter <mreuter(a)nmr.mgh.harvard.edu> > wrote: >> >> Hi Elijah, >> >> if you have minimal edits, I'd recommend to reprocess the data from >> scratch with 5.3 (that is supposed to be the same as using the -clean flag >> when rerunning 5.3 on top of the old 5.0 data). But I'd recommend to create >> a new subjectsdir, copy over the subjedid/mri/orig dirs for each subject and >> rerun from scratch with 5.3 the independent, base and long runs. >> >> Best, Martin >> >> >> >> On 01/13/2014 06:18 PM, Elijah Mak wrote: >> >> Greetings Freesurfer Community, >> >> I am running a longitudinal study using Freesurfer v.5.3 to compare rate >> of cortical thinning between 2 groups of subjects. I hope I can seek some >> expert advice on the following situation: >> >> All data (both the baseline and follow-up) has been processed >> independently using recon-all. This has been done on Freesurfer 5.0. >> >> Since then, I have upgraded to Freesurfer v5.3 to utilize the linear mixed >> effect capability. >> >> For the construction of base and longs, I understand that it is important >> to add -clean flag to my recon-all strings to overwrite brainmask.mgz etc. >> When must this be done? >> >> I am still new to Freesurer, and I'd greatly appreciate any advice >> regarding workflow. >> >> Many Thanks, >> >> Elijah >> >> >> >> -- >> >> Elijah Mak, B.A. >> >> Doctoral Candidate | Psychiatry >> >> University of Cambridge | Department of Psychiatry >> >> Trinity College, Cambridge, CB2 1TQ >> >> >> >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer(a)nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> -- >> Martin Reuter, Ph.D. >> Assistant in Neuroscience - Massachusetts General Hospital >> Instructor in Neurology - Harvard Medical School >> MGH / HMS / MIT >> >> A.A.Martinos Center for Biomedical Imaging >> 149 Thirteenth Street, Suite 2301 >> Charlestown, MA 02129 >> >> Phone: +1-617-724-5652 >> Email: >> mreuter(a)nmr.mgh.harvard.edu >> reuter(a)mit.edu >> Web : http://reuter.mit.edu >> >> The information in this e-mail is intended only for the person to whom it >> is >> addressed. If you believe this e-mail was sent to you in error and the >> e-mail >> contains patient information, please contact the Partners Compliance >> HelpLine at >> http://www.partners.org/complianceline . If the e-mail was sent to you in >> error >> but does not contain patient information, please contact the sender and >> properly >> dispose of the e-mail. > > > > > -- > > Elijah Mak, B.A. > > Doctoral Candidate | Psychiatry > > University of Cambridge | Department of Psychiatry > > Trinity College, Cambridge, CB2 1TQ > > > > -- > Martin Reuter, Ph.D. > Assistant in Neuroscience - Massachusetts General Hospital > Instructor in Neurology - Harvard Medical School > MGH / HMS / MIT > > A.A.Martinos Center for Biomedical Imaging > 149 Thirteenth Street, Suite 2301 > Charlestown, MA 02129 > > Phone: +1-617-724-5652 > Email: > mreuter(a)nmr.mgh.harvard.edu > reuter(a)mit.edu > Web : http://reuter.mit.edu -- Elijah Mak, Gates Scholar PhD Candidate | Psychiatry University of Cambridge | Department of Psychiatry Trinity College, Cambridge, CB2 1TQ

2 1

free-surfer longitudinal Qdec table
by Anna Jonsson 04 Jun '14

04 Jun '14

Hi, I would like to make my freesurfer longitudinal qdec table for the to use in the lme matlab tools. However, does the fsid need to refer to the longitidunally derived subject; eg XXX.long, or the cross-sectionally derived subject, eg XXX. Im not sure based on the specification online. Kind Regards Anna

1 0

v.5.3 Longitudinal Analysis on v.5.0 processed scans
by Elijah Mak 04 Jun '14

04 Jun '14

Greetings Freesurfer Community, I am running a longitudinal study using Freesurfer v.5.3 to compare rate of cortical thinning between 2 groups of subjects. I hope I can seek some expert advice on the following situation: All data (both the baseline and follow-up) has been processed independently using recon-all. This has been done on Freesurfer 5.0. Since then, I have upgraded to Freesurfer v5.3 to utilize the linear mixed effect capability. For the construction of base and longs, I understand that it is important to add -clean flag to my recon-all strings to overwrite brainmask.mgz etc. When must this be done? I am still new to Freesurer, and I'd greatly appreciate any advice regarding workflow. Many Thanks, Elijah -- Elijah Mak, B.A. Doctoral Candidate *|* Psychiatry University of Cambridge *| *Department of Psychiatry Trinity College, Cambridge, CB2 1TQ

2 4

mri_convert ERROR: could not determine file
by Marko Dakovic 04 Jun '14

04 Jun '14

Hi there, I have the same problem as Morgan. During execution of reg-feat2anat the error message appears mri_convert ERROR: could not determine file /usr/local/freesurfer/subjects....If I understood well the solution is in updated version of reg-feat However, I cannot find the updated reg-feat2anat in ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/ <ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/reg-feat2anat> Would you be so kind to upload the file again. Thanks in advance, Best regards, Marko Dakovic, PhD Center for Magnetic Resonance Clinical Center of Serbia Pasterova 2 Belgrade 11000 Serbia

1 0

Re: [Freesurfer] Advice - best method, longitudinal vs. cross-sectional
by Martin Reuter 03 Jun '14

03 Jun '14

Hi Alex, The time distance does not matter. It's best to use a longitudinal method (both for the image processing, to reduce noise, and for the statistics, to gain power). Your design seems to be missing a control group (unless one off your groups is placebo). In your design you can compare changes across the two groups. But if you look at a single group and detect longitudinal change you will not know if it is the drug or something else that caused it. Best Martin Sent via my smartphone, please excuse brevity. -------- Original message -------- From: Alexandru Hanganu <al.hanganu(a)yahoo.ca> Date:06/03/2014 6:15 PM (GMT-05:00) To: FS Mailing List <Freesurfer(a)nmr.mgh.harvard.edu> Subject: [Freesurfer] Advice - best method, longitudinal vs. cross-sectional Hello Everyone, could someone please give us an advice about which method you consider is the best for our study ? we have two groups with MRI at Time 1. Each group received medication. After this we performed another MRI at Time 2 after 2 weeks. The best method for this study is a longitudinal one or a cross-sectional GLM ? We consider that the distance between the time points is too small, and the longitudinal method is not the best choice. Hence, this study should be treated as a cross-sectional one. In this case we think about performing a simple GLM with the contrasts: 0.5 0.5 0.5 0.5 or 1 -1 -1 1 for the groups: 1) grp 1 time 1 2) grp 1 time 2 3) grp 2 time 1 4) grp 2 time 2 we are searching to see whether medication had any impact on the cortical morphology in each group and between the groups. Thank you ! Best regards, Alex.

1 0

?reference(s) supporting DLPFC label
by Fidel Vila 03 Jun '14

03 Jun '14

Dear FS community, I'm looking into DLPFC volume/thickness. Following advice posted before on this forum, I'm adding G_frontal_middle.label + S_frontal_inferior.label + S_frontal_middle.label from Destrieux 2009 atlas as a surrogate for DLPFC. My question is whether folks could provide reference(s) from the literature backing up this approach? Thanks for your time and help, Kev

2 1

TRACULA: Exits with Errors, and ERROR in -qa Flag
by Shantanu Ghosh 03 Jun '14

03 Jun '14

Dear Anastasia, I wrote to the Freesurfer Listserv yesterday about the problem I am facing in running tracula, although my FS is set to version 5.3. However, I debugged some problems in the config file, and now trac-all -prep runs through all the DTI volumes, but exits with errors. Also, it does not recognize the -qa flag. What am I doing wrong? Attached is the log file, and an updated config file. Any help is highly appreciated. Thanks. Shantanu ---------------------------- Original Message ---------------------------- Subject: TRACULA: ERROR in -qa Flag From: "Shantanu Ghosh" <shantanu(a)nmr.mgh.harvard.edu> Date: Mon, June 2, 2014 6:32 pm To: freesurfer(a)nmr.mgh.harvard.edu -------------------------------------------------------------------------- Dear Anastasia and Freesurfer experts, I am trying to set up an analysis of motion parameters using the flag -q from a set of DTI Dicoms located on the Martinos cluster, but I am facing problems -- it says "ERROR: flag -qa unrecognized". My .tcshrc is set to Freesurfer5.3 and FSL. What am I doing wrong? My config file is attached and here are my commands. cd /autofs/space/glia_001/users/shantanu/dtistudy/ setenv PATH /usr/pubsw/packages/python/epd/bin:${PATH} setenv FSENV_KEEP_PROMPT setenv MNE_ROOT /usr/pubsw/packages/mne/stable setenv SUBJECTS_DIR /autofs/space/calvin_001/marvin/1/users/MRI/WMA/recons setenv FSL_DIR /usr/pubsw/packages/fsl/5.0.0 source /usr/local/freesurfer/nmr-stable53-env source /usr/pubsw/packages/mne/stable/bin/mne_setup ----My console output----- glia:shantanu[120] mynmr53 -------- freesurfer-Linux-centos6_x86_64-stable-v5.3.0-20130514 -------- Setting up environment for FreeSurfer/FS-FAST (and FSL) WARNING: /autofs/space/calvin_001/marvin/1/users/MRI/WMA/recons does not exist FREESURFER_HOME /usr/local/freesurfer/stable5_3_0 FSFAST_HOME /usr/local/freesurfer/stable5_3_0/fsfast FSF_OUTPUT_FORMAT nii.gz SUBJECTS_DIR /autofs/space/calvin_001/marvin/1/users/MRI/WMA/recons MNI_DIR /usr/local/freesurfer/stable5_1_0/mni FSL_DIR /usr/pubsw/packages/fsl/5.0.0 MNE software location set to: /usr/pubsw/packages/mne/stable MATLAB software not available /usr/pubsw/packages/mne/stable/bin already in PATH /usr/pubsw/packages/mne/stable/lib added to LD_LIBRARY_PATH /usr/pubsw/packages/mne/stable/share/app-defaults/%N added to XUSERFILESEARCHPATH Note: Remember to set SUBJECTS_DIR and SUBJECT environment variables correctly. Note: FreeSurfer environment is needed to access tkmedit from mne_analyze. glia:shantanu[121] trac-all -qa -s 008801 -c /autofs/space/glia_001/users/shantanu/dtistudy/dticonfigfile.txt ERROR: flag -qa unrecognized -qa -s 008801 -c /autofs/space/glia_001/users/shantanu/dtistudy/dticonfigfile.txt glia:shantanu[122] Thanks in advance for all help. Shantanu -- Shantanu Ghosh, Ph.D. Harvard Medical School & Massachusetts General Hospital Martinos Center for Biomedical Imaging -- Not only is the universe stranger than we think, it is stranger than we can think. -Werner Heisenberg -- Shantanu Ghosh, Ph.D. Harvard Medical School & Massachusetts General Hospital Martinos Center for Biomedical Imaging -- Not only is the universe stranger than we think, it is stranger than we can think. -Werner Heisenberg

1 0

Re: [Freesurfer] Fwd: How to improve/fix TKMEDIT lag
by Vy Dinh 03 Jun '14

03 Jun '14

Hi, I discovered that installing Freesurfer onto my local workstation and running Tkmedit from there has resolved my issue. It doesn't explain why tkmedit is so slow when run from our server, so if you have some ideas please let me know. Thank you all for your help. Best, Vy On Tue, Jun 3, 2014 at 12:39 PM, Vy Dinh <vydinh1(a)gmail.com> wrote: > Thank you for your reply. All of our Freesurfer commands, including > Tkmedit are run from the servers on the RedHat Machine. We use the ssh -Y > syntax (not -X), and when I tried using -X instead, it doesn't change the > speed of tkmedit. > > Additionally, I tried copying my data to another (department's) server at > our institution and running tkmedit from there. We didn't experience any > lag here, so this suggests that our department's dedicated server is > somehow dealing with tkmedit differently. However, when our server expert > checked the CPU usage, it appears that OUR department's server is using > less RAM overall (so it if anything, our data should process more > quickly.). In other words, we can't figure out how Tkmedit is utilized by > the servers differently. > > -Vy > > > On Tue, Jun 3, 2014 at 11:13 AM, dgw <dgwakeman(a)gmail.com> wrote: > >> Hi, >> >> What machine are you actually running tkmedit on? >> >> more specifically: >> You appear to be using a mac. Do you run tkmedit directly on the mac? >> >> or are you using ssh -X (or something like it to run it) on the Red Hat >> machines? >> >> Thanks, >> D >> >> >> On Tue, Jun 3, 2014 at 11:55 AM, Vy Dinh <vdinh2(a)wisc.edu> wrote: >> >>> >>> Greetings Freesurfer developers, >>> >>> I'm looking for ways to improve the speed of TKMEDIT for Freesurfer 5.3. >>> Specifically, the GUI interface is abnormally slow; edits made to a voxel >>> takes 1-2 seconds to show up on the screen. Interesting, the recon-all >>> doesn't appear to be slow (as each subject has been processed within a 24 >>> hr period), so I wonder if this problem has something to do with TKMEDIT's >>> GUI & my current setup. >>> >>> Although I SSH into our server for data processing, here are the specs >>> for my local workstation: >>> >>> Mac OSX 10.9.2 >>> 2.7 GHz Intel Core i5 Determine how many cores (Apple icon >> About this >>> Mac >> More Info) >>> 4 GB 1333 MHz DDR3 >>> >>> Server Specs: >>> >>> Our App servers have 4 Xeon CPUs apiece, each with 8 cores running at >>> 2.7GHz, for a total of 32 cores (I was told that this shouldn't be the >>> cause as the point of multi-core and multi-threading is not dependent on >>> speed only). As well, each server has 94 GB of RAM. >>> >>> Our App servers mount shares via NFS from our file servers. All servers >>> are the same version of Red Hat Linux, so there is no differences >>> there. According to the server expert at our institution, the tkmedit does >>> not appear to surpass a CPU *utilization* of 8*%* of a single core, so >>> the programs running off the server should be able to utilize its maximum >>> speed. >>> >>> I am working closely with our server expert to troubleshoot this issue, >>> so any insight from the development team will be much appreciated. Thank >>> you. >>> >>> >>> Best regards, >>> >>> Vy Dinh >>> MD Candidate, Class of 2017 >>> University of Wisconsin School of Medicine and Public Health >>> >>> _______________________________________________ >>> Freesurfer mailing list >>> Freesurfer(a)nmr.mgh.harvard.edu >>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >>> >>> The information in this e-mail is intended only for the person to whom >>> it is >>> addressed. If you believe this e-mail was sent to you in error and the >>> e-mail >>> contains patient information, please contact the Partners Compliance >>> HelpLine at >>> http://www.partners.org/complianceline . If the e-mail was sent to you >>> in error >>> but does not contain patient information, please contact the sender and >>> properly >>> dispose of the e-mail. >>> >>> >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer(a)nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> The information in this e-mail is intended only for the person to whom it >> is >> addressed. If you believe this e-mail was sent to you in error and the >> e-mail >> contains patient information, please contact the Partners Compliance >> HelpLine at >> http://www.partners.org/complianceline . If the e-mail was sent to you >> in error >> but does not contain patient information, please contact the sender and >> properly >> dispose of the e-mail. >> >> > > > -- > Vy T.U. Dinh > Research Assistant, Neurological Sciences > Rush University Medical Center > Phone: (312) 563-3853 > Fax: (312) 563-4660 > Email: Vy_dinh(a)rush.edu > > -- Vy Dinh MD Candidate, Class of 2017 University of Wisconsin School of Medicine and Public Health

2 3

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 June 2014