Freesurfer June 2016

freesurfer@nmr.mgh.harvard.edu

152 participants
213 discussions

correct thinking?
by Marja Caverlé 02 Jun '16

02 Jun '16

Hi guys, I just wanted to make sure I am understanding everything correct. I have done recon -all -all on all my subjects. Can I now just use the numbers in the .stats files to compare these with my other measures? It seems to be very (almost too) simple, so I that's why I am checking to see whether I am correct. Thanks, Marja

3 6

Selxavg3-sess with no-con-ok and ROI analyses
by Raney, Talia L. 02 Jun '16

02 Jun '16

Hi FreeSurfer users, I have some resting state data that I preprocessed and conducted first level analyses using the flag -no-con-ok. Since no ces.nii.gz or sig.nii.gz is createdcomes up because there is no contrast, I'm wondering if it's possible to conduct functional ROI analyses with the data processed this way. Or would I have to do ROI analyses with the data processed in a seed-based manner where the seed is the "contrast"? Best, Talia

2 4

Postdoctoral position in computational neuroimaging
by Christian Wachinger 02 Jun '16

02 Jun '16

Postdoctoral position in computational neuroimaging The Child and Adolescent Psychiatry at Ludwig Maximilian University (LMU) Munich is looking for a postdoctoral candidate in computational neuroimaging. The researcher will be involved in the development of new algorithms for the analysis of structural and functional magnetic resonance images with a focus on large-scale neuroimaging datasets. The applicant should have: - A Ph.D. degree in computer science, engineering, mathematics, neuroscience or related field - Strong mathematical and problem solving skills - Experience in machine learning and image analysis - Strong programming skills - Experience in neuroimaging is a plus The Ludwig Maximilian University is ranked as the top university in Germany according to Times Higher Education World University Rankings. LMU Munich is an equal opportunity employer. Applications of women and disabled persons - given similar qualification - are preferred. Applicants with children are welcome. Please send your CV, contact details of two academic references and a cover letter to christian.wachinger(a)med.uni-muenchen.de.

1 0

Re: [Freesurfer] Questions about fsgd file & contrast
by jahanvi patel 02 Jun '16

02 Jun '16

Thank you. Is the following setup looks correct? We have 1 group of subjects, half came at TP1 & TP2, half came at TP1 & TP3. And we would like to do a Two Stage longitudinal anaylsis later, is it correct if we put in the following way? *GroupDescriptorFile 1 **Title G2V2 * *Class MaleClass Female* *Variables Age Minutes * *Input TP1_sub1.long.Template1 Female 26 0* *Input TP2_sub1.long.Template1 Female 26 1800* *Input TP1_sub2.long.Template2 Male 20 0* *Input TP3_sub2.long.Template2 Male 20 1000* Thank you Jahanvi > 1) You don't include sex as a covariate. Include it as a class (ie, > Class Male and Class Female rather than Class Main). > 2) You need to include Minutes in your null hypothesis. > 3) Include minutes as a covariate as you have done, then create the > proper contrast matrix. The contrast matrix will depend on the classes > you set up > > You can also see > https://surfer.nmr.mgh.harvard.edu/fswiki/FsgdExamples On 6/1/16 9:03 AM, jahanvi patel wrote: >* Dear list, *>>* We have a question about the set up of the FSGD file. *>>>>* We have 1 group, 2 continued covariates (age & mintues), 1 *>* discontinued covariate (sex) *>>>* GroupDescriptorFile 1 *>* Title G1V3 *>* Class Main *>* Variables Age Sex Minutes *>* Input subject1 Main 30 0 1200 *>* Input subject2 Main 40 1 180 *>>>* My question is *>>>* 1) How to enter the column "sex" in FSGD. *>* Can I input 1 for Male and 0 for Female? *>>* 2) What would be the Main Contrast (regress out effect of sex & age)? *>>* (Null Hypothesis: the intercept/offset is different than 0 after *>* regressing out the effects of age and sex) *>>* 3) If I want to correlate the continued variable (minutes) with the *>* Thickness, is it correct in this way ? : *>>* 0 0 0 1 *>>>>* Thanks so much *>>>* Jahanvi *>>>>* _______________________________________________ *>* Freesurfer mailing list *>* Freesurfer at nmr.mgh.harvard.edu <https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer> *>* https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer <https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer> *

1 0

Group Analysis: Two Groups and Two Covariates.
by Hanbyul Cho 01 Jun '16

01 Jun '16

Dear FreeSurfer Team, I processed the Group analysis by FreeSurfer 5.3.0 Our data has two groups (Patient, Control) and two covariate values (age, extra value) mtx, contrast is 1 -1 0 0 0 0 The mris_preproc, and mri_surf2surf command were completed without error. mris_preproc --fsgd **G2V2.fsgd --target fsaverage --hemi lh --meas thickness --out lh.**G2V2.dods.mgh mri_surf2surf --hemi lh --s fsaverage --sval lh.**G2V2.dods.mgh --fwhm 15 --cortex --tval lh.**G2V2.dods_sm15.mgh mri_glmfit --y lh.**G2V2.dods_sm15.mgh --fsgd **.fsgd dods --C **G2V2.mtx --surf fsaverage lh --cortex --glmdir lh.**G2V2.dods_sm15.glmdir After mri_glmfit command, I saw the error message. ========================================================= ERROR: matrix is ill-conditioned or badly scaled, condno = 1e+08 -------------------------------- Possible problem with experimental design: Check for duplicate entries and/or lack of range of continuous variables within a class. If you seek help with this problem, make sure to send: 1. Your command line: mri_glmfit --y lh.**G2V2.dods_sm15.mgh --fsgd **.fsgd dods --C **G2V2.mtx --surf fsaverage lh --cortex --glmdir lh.**G2V2.dods_sm15.glmdir 2. The FSGD file (if using one) 3. And the design matrix above ========================================================= I think our fsgd and mtx compositions are equal to the g2v2.fsgd which FreeSurfer wiki example. (https://surfer.nmr.mgh.harvard.edu/fswiki/Fsgdf2G2V) In this situation, how can I correct my fsgd or mtx files? Is there any other command option? Best Regards, Han

2 9

Masking in functional connectivity pipeline
by Rodriguez-Thompson, Anais 01 Jun '16

01 Jun '16

Hi FreeSurfer experts, I have a question about two specific steps in my fc pipeline. Since my seed is a single vertex, I convert the seed label to a mask, which I then smooth and binarize, and then convert to a volume file for fcseed-sess. 1) Is there way to directly convert a label into a cortical mask? Right now I use mri_label2label to create an outmask mri_label2label -s fsaverage -regmethod -surface -hemi -srclabel $SUBJECTS_DIR/{subject}/label/{label} -trglabel $SUBJECTS_DIR/{subject}/label/junk.label -outmask $SUBJECTS_DIR/{subject}/mri/{mask} cd $SUBJECTS_DIR/{subject}/label rm junk.label Is there a way to do this where I do not have to create another label in order to make a mask? 2) I can't figure out how to convert a cortical mask to a volume file (that I can use in fcseed-sess) using stable 5.3. I've tried to use mri_convert to create a volume file, but fcseed-sess errors out using this vol file (0 segmentations, core segmentation dumped) The command I use is: mri_convert $SUBJECTS_DIR/{subject}/mri/{mask} $SUBJECTS_DIR/{subject}/mri/{mask}.vol.mgz Currently, if I want to do this I need to switch to stable dev and run mri_surf2vol: mri_surf2vol -so $SUBJECTS_DIR/fsaverage1/surf/lh.pial $SUBJECTS_DIR/{subject}/mri/{mask} -subject fsaverage -o $SUBJECTS_DIR/{subject}/mri/{mask}.vol.mgz Is there a way that I can convert a surface mask to a useable volume for fcseed-sess in stable 5.3, rather than having to switch to stable dev? Thank you for your help! Anais Rodriguez-Thompson Clinical Research Coordinator Brain Genomics Laboratory Martinos Center for Biomedical Imaging Massachusetts General Hospital Phone: (617) 643-3215

1 0

Re: [Freesurfer] Using customized stats code for mass-univariate approach in freesurfer for structural data
by Chung, Yoonho 01 Jun '16

01 Jun '16

Yes - I am interested in applying the monte carlo sim to my own model fit. Doug can you walk me through the steps I need to get this part done? When you say residuals, I am assuming that I need to supply a vector file of residuals after I fit my model to the smoothed data (eg. using mri_surf2surf --sval [concat thickness] --fwhm-trg [smooth level] etc. ) - correct? Since I don't have the --glmdir to supply, what are the other files or options that I need to manually supply for mri_glmfit-sim to run the correction? Thank you in advance!

2 2

FC from resting-state
by Martin Juneja 01 Jun '16

01 Jun '16

Hello FreeSurfer experts, I am working on resting-state functional connectivity using freesurfer by following the instructions from: http://surfer.nmr.mgh.harvard.edu/fswiki/FsFastFunctionalConnectivityWalkth… I made sure that the seed region (for ever subject) I am using is at the correct place on raw structural image. After running the analysis and looking at the RSFC maps, very few subjects showed very high FC at or close to seed region but for most of the subjects I am not getting high FC at/close to seed region. I would really appreciate any suggestion or tips on this e.g. what could be wrong during the analysis? Thanks. MJ

2 1

mris_preproc - srcsubjreg
by Makaretz, Sara Johanna 01 Jun '16

01 Jun '16

Hi Doug, I¹m curious why the newer mris_preproc does not use sphere.reg as the default srcsurfreg when the target is not fsaverage - see below Thanks for any info! (not urgent) On 2016-5-24, 11:30, "Z K" <zkaufman(a)nmr.mgh.harvard.edu> wrote: >Good question, and I dont know the answer. mris_preproc is Dougs baby so >I suggest just posting this question to the list with "mris_preproc" in >the subject, and I bet he will give you the answer. > >-Zeke > >On 05/23/2016 10:31 PM, Makaretz, Sara Johanna wrote: >> Hi Zeke, >> >> I am running some surface glms where targ is a patient surface. The help >> says --target TargetSubject (requires ?h.TargetSubject.sphere.reg), and >> I was getting errors because dev-default mris_preproc wanted >> mri_surf2surf to have srcsurfreg = >> $SUBJECTS_DIR/$srcsubject/surf/$h.$TARGSUBJ.sphere.reg >> >> But mris_preproc in 5.3 let me do target = patient surface with no extra >> surf reg files or extra flags to get mri_surf2surf to use srcsurfreg = >> $SUBJECTS_DIR/$srcsubject/surf/$h.sphere.reg >> >> I had been running this with 5.3: >> mris_preproc \ >> --target $TARGSUBJ \ >> --hemi lh \ >> --meas thickness \ >> --fsgd file.fsgd \ >> --out lh.output.mgh >> >> And I was about to email freesurfer@nmr but "for fun" tried adding >> --surfreg sphere.reg which got mris_surf2surf to use the right reg >> files. So now I'm selfishly trying to figure this out - do you know what >> prompted the switch from sphere.reg "opt-out" to sphere.reg "opt-in"? >> Aka why sphere.reg is non-default when the target is non-fsaverage? Did >> that make sense? I know there must be some reason, so I really want to >> know a bit more so I can stay sane when old scripts break. >> >> Thanks for any info! Sorry to bother you! >> Sara >> >> >> >> >> >> >> Outputs from the FS-dev one: >> >> $Id: mris_preproc,v 1.76 2015/12/09 20:02:53 zkaufman Exp $ >> .... >> Src lh $TARGSUBJ.sphere.reg >> Trg lh sphere.reg >> .... >> mri_surf2surf --srcsubject $SRCSUBJ --srchemi lh --srcsurfreg >> $TARGSUBJ.sphere.reg --trgsubject $TARGSUBJ --trghemi lh --trgsurfreg >> sphere.reg --etcetcetc >> .... >> ERROR can't find $SUBJECTS_DIR/$SRCSUBJ/surf/$TARGSUBJ.sphere.reg >> because I never made it >> >> >> >> Outputs from the FS-dev one with fsaverage as the targ subj: >> >> $Id: mris_preproc,v 1.76 2015/12/09 20:02:53 zkaufman Exp $ >> .... >> Src lh sphere.reg >> Trg lh sphere.reg >> .... >> mri_surf2surf --srcsubject $SRCSUBJ --srchemi lh --srcsurfreg sphere.reg >> --trgsubject fsaverage --trghemi lh --trgsurfreg sphere.reg --etcetcetc >> .... >> mris_preproc finds everything, runs no problem >> >> >> >> Outputs from the FS-5.3 one: >> >> $Id: mris_preproc,v 1.59.2.4 2012/12/06 16:06:52 mreuter Exp $ >> .... >> Src lh sphere.reg >> Trg lh sphere.reg >> .... >> mri_surf2surf --srcsubject $SRCSUBJ --srchemi lh --srcsurfreg sphere.reg >> --trgsubject $TARGSUBJ --trghemi lh --trgsurfreg sphere.reg --etcetcetc >> .... >> mris_preproc exits with no errors because everything exists >> >> >> >>

4 5

Questions about fsgd file & contrast
by jahanvi patel 01 Jun '16

01 Jun '16

Dear list, We have a question about the set up of the FSGD file. We have 1 group, 2 continued covariates (age & mintues), 1 discontinued covariate (sex) GroupDescriptorFile 1 Title G1V3 Class Main Variables Age Sex Minutes Input subject1 Main 30 0 1200 Input subject2 Main 40 1 180 My question is 1) How to enter the column "sex" in FSGD. Can I input 1 for Male and 0 for Female? 2) What would be the Main Contrast (regress out effect of sex & age)? (Null Hypothesis: the intercept/offset is different than 0 after regressing out the effects of age and sex) 3) If I want to correlate the continued variable (minutes) with the Thickness, is it correct in this way ? : 0 0 0 1 Thanks so much Jahanvi

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 June 2016