Freesurfer

freesurfer@nmr.mgh.harvard.edu

19 participants
30132 discussions

Surface-based inter-subject registration of both anatomical and functional images
by Joonkoo Park 17 Feb '10

17 Feb '10

Hi, I'm completely new to FreeSurfer, and I'd like to get some guide to surface-based inter-subject registration in analyzing some fMRI data. I have an anatomical image (e.g. t1-spgr or mprage) and functional images (e.g. epi) in NIFTI or ANALYZE format for each subject. In each subject's native space (typically using SPM), I perform some GLM and get some statistical parametric maps (SPMs) for each subject (note that the anatomical image and the resulting functional image are co-registered). Now, I'd like to use the surface-based normalization (or I realized that this procedure is called registration in the freesurfer community) procedure to normalize the anatomical and the functional images to a standard space. Could you please guide me what steps I need to perform in order to achieve this goal? Here are some specific questions: 1. What command shall I use to get the nifti/analyze images into the freesurfer format? I was wondering if recon-all is really necessary. I heard that it takes more than a day to do one subject, and I don't think I need all those results for my purpose. 2. I found that the command "mris_register" will give me what I want at least for the anatomical data. But how do I do the same for the functionals? Also, how will I make sure that my anatomicals and functionals are co-registered? 3. I need to be able to get into the data structure after inter-subject normalization. For the nifti/analyze files, there's an SPM command that will bring all the data into a matrix in Matlab. But for the freesurfer format, is there a way to import the surface-based files into Matlab so that I can play with the output data? I may not be asking in the right order or even in the right way. But I'm thinking that any FreeSurfer guru should be able to help me. I searched through the archive and the documentations but wasn't satisfactory. Thanks in advance! Joon -- Joonkoo Park Ph.D. Candidate, Cognition & Cognitive Neuroscience Department of Psychology, University of Michigan

3 2

difference between structures...
by Gonzalo Rojas Costa 17 Feb '10

17 Feb '10

Hi: There exist any tool that could show me the structures that differ between each hemisphere more than a given percentage (volume) ?... For example: the hippocampus... Sincerely, Gonzalo Rojas Costa

2 1

question about visualization of negative activation
by Balaji Goparaju 17 Feb '10

17 Feb '10

Hello, I have a loaded a significance map using the tksurfer-sess command. I want to visualize only the negative activation on the surface. In the View -> Configure -> Overlay window, I check both the "truncate" and "inverse" buttons. So only the negative activation is shown however it is in Red-Orange-Yellow color scale. How can I get this negative activation to be represented by Dark and Light Blue color scale. Thanks in advance for the help! Balaji Goparaju

2 1

Bbregister
by Lena Palaniyappan 17 Feb '10

17 Feb '10

I want to use contrast files that are outputs from SPM (in IMG HDR format) to overlay on inflated surfaces. But if I use bbregister and choose T1 as the image type, the regisration achieved is poor (image attached) Is this the correct way of doing this or am I missing something? Commands used: bbregister --mov con_0001.img --t1 --s fsaverage --init-spm --reg register.dat cat register.dat tkregister2 --mov con_001.img --reg register.dat --surf Many thanks Lena This message has been checked for viruses but the contents of an attachment may still contain software viruses which could damage your computer system: you are advised to perform your own checks. Email communications with the University of Nottingham may be monitored as permitted by UK legislation.

4 6

inter-subject mri_vol2surf from SPM8 functional volume data
by Nela Cicmil 17 Feb '10

17 Feb '10

Dear Freesurfer users, Thank you for your previous help and advice. I am still having difficulty creating reasonable-looking surface overlay files from the functional data in volume format, and would be grateful for any further help. I have a functional image in volume format (originally written to .nii format, converted with fslchfiletype ANALYZE to .img format), which is a SPM image of MEG activity in particular subject subject1, created with SPM8 MEG analysis software (UCL). I would like to display this activity as a surface overlay on the cortical surface of a different subject, subject2, in tksurfer. I use tkregister2, with --mov as the functional image .img and --s as subject2, to align the two images and create an appropriate register.dat file. When I try to create a surface overlay from the functional image: mri_vol2surf --mov f.img --out f-rh.img --reg register.dat --hemi rh --surf inflated - the resulting suface overlay does not correspond well to the activation seen on the original volume image. For example, in the functional volume image viewed in tkmedit on subject1's anatomy, activity which is around the occipital pole (early visual areas) when converted through this method to a surface overlay is now very much anteriorly shifted (i.e occipital-temporal junction). It is the same whether I use .w or .mgh as output file types. I have tried a method avoiding the register.dat file: mri_vol2surf --mov f.img --out f-rh.img --regheader subject1 --srcsubject subject1 --trgsubject subject2 --hemi rh --surf inflated - the activity pattern on the surface overlay is almost identical to previous attempt and registration is not improved. To check whether this was just an inter-subject problem, I created new register.dat file to map the functional volume onto subject1's own cortical surface and repeated mri_vol2surf with --reg register.dat. The activity pattern on the new cortical surface is still very anteriorly shifted in the same way as in the previous attempt in the other subject. Could there be some kind of coordinate mis-alignment between my functional image and the anatomical surfaces? As a novice to these kinds of image manipulations, I can't guess at the solution! Thanks for your help, Nela

2 1

mri_vol2surf error
by Gregory Dierksen 17 Feb '10

17 Feb '10

Hi, I'm trying to paint functional data onto a cortical "midgray" surface. The surface was generated using mris_expand to expand the WM surface to 50% of the pial surface. The data I want to paint onto this surface is in a 96 x 96 x 17 x 512 dimension NIFTI file generated using the feat command (the volume is the resulting "filtered_func_data.nii.gz"). When using this command: mri_vol2surf --mov filtered_func_data.nii.gz \ --reg register.dat \ --hemi rh \ --surf midgray.05 \ --nvox NonZero_voxels.txt \ --out Total_TimeSeries_RHsurface.nii.gz I get the following error message: NIFTI FORMAT WARNING: ncols 65938 in input exceeds 32768. So I'm going to put the true ncols in glmin and set dim[1]=-1. This should be ok within FreeSurfer, but you will not be able to use this volume with other software. And an output volume with dimensions 65938 x 1 x 2 x 512 which does not work with FSL commands. This is a problem since the next thing I want to do is to use fslmaths to collapse the surface in the X, Y, and Z dimensions to obtain a mean for the surface. Also confusing is the fact that the above command works fine for the left hemisphere. What's going on? Am I doing something wrong?

2 1

Re: [Freesurfer] Wmparc Destrieux atlas
by Lena Palaniyappan 17 Feb '10

17 Feb '10

aparc2aseg using Destrieux atlas now seems to run OK for a different subject anyway - and I managed to get the same image as you did. But there are two problems: 1. when I do tkmedit subjid T1.mgz -segmentation wmparc.a2009s.mgz with color LUT, it still shows Killiany/Desikan conventions for wm parcels. Why is this? 2. despite the above trouble I ran mri_segstats --seg MPRAGE07/mri/wmparc.a2009s.mgz --sum MPRAGE07/stats/wmparc.a2009s.stats --pv MPRAGE07/mri/norm.mgz --excludeid 0 --brain-vol-from-seg --brainmask MPRAGE07/mri/brainmask.mgz --in MPRAGE07/mri/norm.mgz --in-intensity-name norm --in-intensity-units MR --etiv --subject MPRAGE07 --surf-wm-vol --ctab $FREESURFER_HOME/FreeSurferColorLUT.txt (MPRAGE07 is my subjid) - it ended prematurely declaring atlas_icv (eTIV) = 1299492 mm^3 (det: 1.499128 ) Loading MPRAGE07/mri/wmparc.a2009s.mgz Getting Cerebral WM volumes from surface lh white matter volume 210579 rh white matter volume 206282 Loading MPRAGE07/mri/norm.mgz Loading MPRAGE07/mri/norm.mgz Loading MPRAGE07/mri/brainmask.mgz # nbrainmaskvoxels 1469167 # brainmaskvolume 1469167.0 # nbrainsegvoxels 252233 # brainsegvolume 252233.0 Voxel Volume is 1 mm^3 Generating list of segmentation ids Found 1190 segmentations Computing statistics for each segmentation 1 1 Left-Cerebral-Exterior 0 2 2 Left-Cerebral-White-Matter Segmentation fault Why is this? Many thanks Lena On 16/02/2010 19:27, "Nick Schmansky" <nicks(a)nmr.mgh.harvard.edu> wrote: > which version of freesurfer is being used? (although i dont think it > should matter). > > is the talairach.xfm correct? > > attached is a snapshot of running what i suggested on a subject of my > own, and it appears correct. > > i am using this to display: > > tkmedit subjid T1.mgz -segmentation wmparc.mgz > > which uses the default color table. > > n. > > On Tue, 2010-02-16 at 18:21 +0000, Lena Palaniyappan wrote: >> Dear Nick >> >> This is exactly what I have been doing for last 3 days!!! - took the last 2 >> commands and made changes as you suggested. >> But the resultant wmparc.a2009s.mgz is bizarre - showing lots of out of >> bounds regions and pointing parahippocampus somewhere near frontal cortex >> (when using LUT to view the segmentation) - see attached >> >> Is this a problem with LookupTable? >> I am not sure if I will get the statistics if I move on to segstats with >> this issue >> (I was about to giveup when this reply came in-- THANKS!) >> >> Lena >> >> >> >> On 16/02/2010 17:33, "Nick Schmansky" <nicks(a)nmr.mgh.harvard.edu> wrote: >> >>> To get the Destrieux atlas parcellations in the wmparc, first have a >>> look at the -wmparc section of the recon-all table: >>> >>> http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable >>> >>> those last two commands would be modified like this: >>> >>> mri_aparc2aseg --s <subjid> --labelwm --hypo-as-wm --rip-unknown >>> --ribbon --o mri/wmparc.a2009s.mgz --ctxseg aparc.a2009s+aseg.mgz >>> >>> >>> mri_segstats --seg mri/wmparc.a2009s.mgz --sum >>> stats/wmparc.a2009s.stats ... (all those other args) >>> >>> >>> the important change was to mri_aparc2aseg, where i change the --ctxseg >>> arg to aparc.a2009s+aseg.mgz, as that is the Destrieux atlas. the other >>> change was just changing the 'wmparc' name to 'wmparc.a2009s', just to >>> be consistent with naming. >>> >>> probably we should modify recon-all so that wmparc.a2009s.stats is >>> created by default. >>> >>> you could also 'cheat' by just copying aparc.a2009s+aseg.mgz overtop >>> aparc+aseg.mgz and then running 'recon-all -s subj -wmparc', but that >>> can make keeping track of what's what more difficult. >>> >>> n. >>> >>> >>> >>> On Mon, 2010-02-15 at 15:19 +0000, Lena Palaniyappan wrote: >>>> >>>> Hi all >>>> >>>> Could anyone help me with extracting the wmparc volume stats in >>>> accordance with Destrieux atlas? >>>> >>>> As it stands now the segmentations corresponding to whitematter >>>> beneath Destrieux 2009 parcellations are returned empty when called >>>> for >>>> >>>> Many thanks >>>> Lena >>>> >>>> \_________________________________ >>>> Lena Palaniyappan >>>> Clinical Lecturer & Honorary StR | Division of Psychiatry ( University >>>> of Nottingham) >>>> South Block 'A' floor| Queens Medical Centre | Nottingham | NG7 2UH >>>> >>>> >>>> >>>> >>>> This message has been checked for viruses but the contents of an >>>> attachment may still contain software viruses which could damage your >>>> computer system: you are advised to perform your own checks. Email >>>> communications with the University of Nottingham may be monitored as >>>> permitted by UK legislation. >>>> >>>> _______________________________________________ >>>> Freesurfer mailing list >>>> Freesurfer(a)nmr.mgh.harvard.edu >>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>> >> >> _______________________________________________________ >> Lena Palaniyappan >> Clinical Lecturer & Honorary StR | Division of Psychiatry ( University of >> Nottingham) >> South Block 'A' floor| Queens Medical Centre | Nottingham | NG7 2UH >> >> _______________________________________________________ Lena Palaniyappan Clinical Lecturer & Honorary StR | Division of Psychiatry ( University of Nottingham) South Block 'A' floor| Queens Medical Centre | Nottingham | NG7 2UH

1 0

ROI based on Talairach coordinate
by Ben Hamida Cammoun Leila Faiza 16 Feb '10

16 Feb '10

Dear Freesurfer users and experts, This is my second attempt :( I am trying to construct a roi based on some talairach coordinate coming from a paper, but I completely fail to do it with matlab and neither manually with tksurfer. My first try was to use the transformation Talairach.xfm with matlab, inverse it and try to recover RAS coordinate, it gives me completely wrong values. Second try was to introduce the Tal coordinate directly in Tksurfer, which is not possible I think (correct me if I am wrong), with the version I have the is a bug in tkmedit I guess when introducing Talairach coord (the ideas was to save them and to load the vertex in tksurfer).. the values changes in the edit box.. Is there another way?? Is it possible with script? Or how to do this by matlab? Please don't ignore the request. Leila

2 3

tractography and freesurfer
by Dana W. Moore 16 Feb '10

16 Feb '10

I have a colleague who is interested in the DTI toolkit to coregister DTI results with FreeSurfer's segmentation. Is that available in any form yet? Thanks, Dana Dana W. Moore, Ph.D. Neuropsychology Fellow Cornell Neuropsychology Service Weill Medical College of Cornell University New York Presbyterian Hospital Department of Neurology & Neuroscience 428 East 72nd Street, Suite 500 New York, NY 10021 Phone: 212-746-2441 Fax: 212-746-5584 Email: dwm2003(a)med.cornell.edu

2 1

Gyrification index for individual regions
by Lena Palaniyappan 16 Feb '10

16 Feb '10

I managed to obtain outputs from recon-all lGI now I would like to obtain a mean gyrification index (not sure if the mean is meaningful?!) for all vertices within 3 ROIs as defined by Destrieux atlas - How should I do this? Can I treat lGI surfaces just as thockness maps in which case can I use mri_segstats to obtain mean values? Cheers Lena This message has been checked for viruses but the contents of an attachment may still contain software viruses which could damage your computer system: you are advised to perform your own checks. Email communications with the University of Nottingham may be monitored as permitted by UK legislation.

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