External Email - Use Caution
Hi Bruce and Doug,
To respond to your questions:
I projected the BOLD data on the surface using mri_vol2surf function. Then
obtained a .mgz file of size 10242 x 1 x 1 x 200.
Then I performed diffusion embedding (unsupervised dimensionality reduction
technique) on this .mgz file, whose output was a .npy file.
I then loaded this file into a matrix on python, which size was 1 x 18715 x
5.
I then saved this file as a .mat file.
Then loaded it on matlab, and kept only 1 x 18715 x 1 part of the matrix,
and stored it into a zeros(1, 20484).
Then kept only the part from (0:10242) and used save_mgh function to save
this data into a .mgz file
Looking at the info of my .mgz file created with save_mgh, I see that the
file doesn't seem to include the data:
"
mri_info lh_emb_sub-002_compassion.mgz
Volume information for lh_emb_sub-002_compassion.mgz
type: MGH
dimensions: 1 x 16 x 1
voxel sizes: -nan, -nan, -nan
type: FLOAT (3)
fov: -nan
dof: 1
xstart: -nan, xend: -nan
ystart: -nan, yend: -nan
zstart: -nan, zend: -nan
TR: 0.00 msec, TE: 0.00 msec, TI: 0.00 msec, flip angle: 0.00
degrees
nframes: 1
PhEncDir: UNKNOWN
FieldStrength: 0.000000
ras xform present
xform info: x_r = -nan, y_r = -nan, z_r = -nan, c_r =
-nan
: x_a = -nan, y_a = -nan, z_a = -nan, c_a =
-nan
: x_s = -nan, y_s = -nan, z_s = -nan, c_s =
-nan
talairach xfm :
Orientation : III
Primary Slice Direction: axial
voxel to ras transform:
-nan -nan -nan -nan
-nan -nan -nan -nan
-nan -nan -nan -nan
0.0000 0.0000 0.0000 1.0000
voxel-to-ras determinant -nan
ras to voxel transform:
mat = NULL!
MatrixFree: NULL mat POINTER!
"
It seems that save_mgh function requires tclsh package, as when I ran
save_mgh for the first time with this command line:
save_mgh('freeview_overlay',
['/home/loic/Documents/emb_output/mgh_format/emb_' char(subjectsList(1))
'_' char(statesList(1)) '.MGZ'], M);
I was getting this message in return in the matlab window:
"
/usr/share/modules/init/bash: line 36: /usr/bin/tclsh: No such file or
directory
/usr/share/modules/init/bash: line 58: export: _moduleraw: not a function
/usr/share/modules/init/bash: line 60: export: module: not a function
"
Checking into the '/usr/bin/tclsh' directory, the only package
called/related to tclsh was 'tclsh8.6'.
I thus changed the name from 'tclsh8.6' to 'tclsh'.
Then tried to save the matrix again with save_mgh function. Now I don't see
the error message anymore.
However I still get a .mgz file with nans, as seen in the previous mri_info
message above.
Would any of you have an advice on how I could make this work, in order to
visualize this .mgz file on the left hemisphere pial surface in Freeview?
Thanks in advance,
Sincerely,
Loïc
Le ven. 21 juin 2019 à 16:24, Bruce Fischl <fis...@nmr.mgh.harvard.edu> a
écrit :
> I don't think it matters but I would use eye instead of nan for M. But the
> critical issue is where you get your data and what dimensions it has
> cheers
> Bruce
>
> On Fri, 21 Jun 2019, Loïc Daumail wrote:
>
> >
> > External Email - Use Caution
> >
> > Me again,
> >
> > Also, here is how I save the matrix on matlab:
> >
> > M = nan(4,4);
> > save_mgh('freeview_overlay',
> ['/home/loic/Documents/emb_output/mgh_format/lh_emb_'
> > char(subjectsList(1)) '_' char(statesList(1)) '.mgz'], M);
> >
> > Thanks,
> >
> > Sincerely,
> >
> > Loïc
> >
> > Le ven. 21 juin 2019 à 12:25, Loïc Daumail <loic.daum...@gmail.com> a
> écrit :
> > Hi Bruce,
> >
> > Thanks for your tip. Unfortunately, I am still running through an
> error...
> > I have a matrix of 18715 elements regarding both hemispheres.
> > The size of fsaverage5 is 20484 I believe, so I filled the remaining
> elements with zeros, and
> > also tried with rand().
> >
> > The other issue is that I am not so sure which part of the matrix is for
> lh and the other one
> > for rh. Thus, I divided 18715/2 and took the 9357 first elements of the
> 18715 matrix. Then
> > filled it with zeros or random values to reach 10242.
> >
> > Then I saved it with save_mgh function and tried to load it with
> freeview using the following
> > command line:
> > freeview
> -f~/Documents/recon_all_success/freesurfer/fsaverage5/surf/lh.pial:overlay=~/Documents/emb_output/mgh
> > _format/lh_emb_sub-002_compassion.mgz
> >
> > I am getting the following error:
> > "
> > Did not find any volume info
> > 2 [0]PETSC ERROR:
> ------------------------------------------------------------------------
> > [0]PETSC ERROR: Caught signal number 11 SEGV: Segmentation Violation,
> probably memory access
> > out of range
> > [0]PETSC ERROR: Try option -start_in_debugger or
> -on_error_attach_debugger
> > [0]PETSC ERROR: or see
> >
> http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html#Signal[0]PETSC
> ERROR:
> > or try http://valgrind.org on linux or man libgmalloc on Apple to find
> memory corruption
> > errors
> > [0]PETSC ERROR: configure using --with-debugging=yes, recompile, link,
> and run
> > [0]PETSC ERROR: to get more information on the crash.
> > [0]PETSC ERROR: --------------------- Error Message
> ------------------------------------
> > [0]PETSC ERROR: Signal received!
> > [0]PETSC ERROR:
> ------------------------------------------------------------------------
> > [0]PETSC ERROR: Petsc Release Version 2.3.3, Patch 13, Thu May 15
> 17:29:26 CDT 2008 HG
> > revision: 4466c6289a0922df26e20626fd4a0b4dd03c8124
> > [0]PETSC ERROR: See docs/changes/index.html for recent updates.
> > [0]PETSC ERROR: See docs/faq.html for hints about trouble shooting.
> > [0]PETSC ERROR: See docs/index.html for manual pages.
> > [0]PETSC ERROR:
> ------------------------------------------------------------------------
> > [0]PETSC ERROR: Unknown Name on a linux-gnu named loic-HP-ZBook-17 by
> loic Fri Jun 21 12:15:53
> > 2019
> > [0]PETSC ERROR: Libraries linked
> from/autofs/space/lyon_006/pubsw/Linux2-2.3-x86_64/packages/petsc/2.3.3-p13/src/petsc-2.3.3-p13/lib/lin
> > ux-gnu-c-opt
> > [0]PETSC ERROR: Configure run at Tue Aug 10 15:01:59 2010
> > [0]PETSC ERROR: Configure options --with-debugging=no --with-cc=gcc
> --with-fc=g77
> > --download-f-blas-lapack=0 --download-mpich=1 --with-mpi=1 --with-x=0
> > --with-gnu-copyright-code=0 --with-shared=0 COPTFLAGS=-O3
> CXXOPTFLAGS=-O3 FOPTFLAGS=-O3
> > [0]PETSC ERROR:
> ------------------------------------------------------------------------
> > [0]PETSC ERROR: User provided function() line 0 in unknown directory
> unknown file
> > [unset]: aborting job:
> > application called MPI_Abort(MPI_COMM_WORLD, 59) - process 0"
> >
> > Any idea of what could be wrong in my process?
> >
> > Thanks
> >
> > Sincerely,
> >
> > Loïc
> >
> >
> >
> > Le jeu. 20 juin 2019 à 19:39, Bruce Fischl <fis...@nmr.mgh.harvard.edu>
> a écrit :
> > Hi Loïc
> >
> > the .mgz file should having dimensions nvertices x 1 x 1 for
> freeview to
> > know it is a surface overlay I believe
> >
> > cheers
> > Bruce
> >
> >
> >
> > On Thu, 20 Jun 2019, Loïc Daumail wrote:
> >
> > >
> > > External Email - Use Caution
> > >
> > > Hi Bruce,
> > >
> > > I am reaching out to you, as I am trying to load the surface
> scalar values of
> > diffusion embedding
> > > for overlay on a pial surface in Freeview.
> > >
> > > I used save_mgh to save a scalar matrix containing surface
> embedding values in
> > an .mgz format.
> > > However I can't load it in freeview, be either only half of the
> matrix data (one
> > hemisphere I
> > > assume) or the whole matrix saved in the same file.
> > >
> > > Is there anything specific I need to do in order to be able to
> visualize a
> > scalar overlay on a
> > > surface? Is .mgz a wrong format for scalar values? I also tried
> with .nii.gz, or
> > .mgh. I am not even
> > > sure save_mgh can do all this...
> > >
> > > Thanks in advance,
> > >
> > > Sincerely,
> > >
> > > Loïc
> > >
> > >
> > >
> > > Le lun. 27 mai 2019 à 11:54, Loïc Daumail <
> loic.daum...@gmail.com> a écrit :
> > > Hi Tim,
> > >
> > > Your link works perfectly, it's awesome!
> > > Thanks a lot,
> > >
> > > Best,
> > >
> > > Loïc
> > >
> > > Le lun. 27 mai 2019 à 11:43, Loïc Daumail <
> loic.daum...@gmail.com> a écrit :
> > > Hi Tim,
> > >
> > > 'mean_rest' is a diffusion embedding map, which contains scores
> for every vertex
> > on
> > > fsaverage5 surface resolution.
> > > 'surf.b' is the pial surface of both hemispheres of
> fsaverage5... I also tried
> > forjust
> > > lh hemisphere, but it gives me the same result...
> > >
> > > Thanks for the link!
> > >
> > > Best,
> > >
> > > Loïc
> > >
> > > Le lun. 27 mai 2019 à 11:20, Tim Schäfer <ts...@rcmd.org> a
> écrit :
> > > External Email - Use Caution
> > >
> > > Hi Loïc,
> > >
> > > that does not look correct, the data is definitely off.
> I'm not sure where
> > > the variables `mean_rest` and `surf.b` come from in your
> code, and I
> > suspect
> > > that something is wrong with them.
> > >
> > > In the meantime, I got interested in this myself and put
> together some
> > code
> > > that will load the Desikan data for both hemispheres and
> display it in
> > > surfstat. Feel free to use it, there also is an example
> screenshot of the
> > > results:
> > >
> > >
> > >
> >
> https://github.com/dfsp-spirit/comp_neuro_science/tree/master/surfstat_desikan
> > >
> > > Best,
> > >
> > > Tim
> > >
> > > --
> > > Dr. Tim Schäfer
> > > Postdoc Computational Neuroimaging
> > > Department of Child and Adolescent Psychiatry,
> Psychosomatics and
> > > Psychotherapy
> > > University Hospital Frankfurt, Goethe University Frankfurt
> am Main,
> > Germany
> > >
> > > > On May 27, 2019 at 11:01 AM Loïc Daumail <
> loic.daum...@gmail.com> wrote:
> > > >
> > > >
> > > > External Email - Use Caution
> > > >
> > > > Hi Tim,
> > > >
> > > > Thanks a lot for your help, indeed, this is exactly what
> I needed!
> > > > I wrote this :
> > > > "
> > > > subjects_dir =
> '/home/loic/Documents/recon_all_success/freesurfer/';
> > > > subject_id = 'sub-004';
> > > > hemi = 'lh';
> > > >
> > > > aparc_file_this_hemi = fullfile(subjects_dir,
> subject_id, 'label',
> > > > sprintf("%s.aparc.annot", hemi));
> > > > [vertices, label, colortable] =
> read_annotation(aparc_file_this_hemi);
> > > > fprintf("Read aparc parcellation file '%s' containing %d
> regions.\n",
> > > > aparc_file_this_hemi, length(colortable.struct_names));
> > > > for sidx = 1:length(colortable.struct_names)
> > > > region = colortable.struct_names{sidx};
> > > > struct_code = colortable.table(sidx, 5);
> > > > vertices_of_struct_roi = find(label == struct_code);
> > > > fprintf("Found region '%s' with %d verts.\n", region,
> > > > length(vertices_of_struct_roi));
> > > > end
> > > > fig1 = figure();
> > > > SurfStatView(mean_rest, surf.b);
> > > > SurfStatColormap(colortable.table(:,1:3)/255);
> > > > "
> > > > However, the image I obtain seems a bit fuzzy...( see
> attached)
> > > > Is it because of my data?
> > > >
> > > > Thanks in advance,
> > > >
> > > > Loïc
> > > >
> > > >
> > > > Le jeu. 23 mai 2019 à 17:20, Tim Schäfer <ts...@rcmd.org>
> a écrit :
> > > >
> > > > > External Email - Use Caution
> > > > >
> > > > > Afaik SurfStat is not maintained anymore and I think
> it's been this
> > way
> > > > > for quite a while now. Maybe someone else knows better?
> > > > >
> > > > > The website is at
> http://www.math.mcgill.ca/keith/surfstat/ and it
> > says
> > > > > 'Updated 26 September 2008'. I'm not sure whether
> that's really true,
> > > but
> > > > > at least for the last 14 months (the only time frame I
> know about),
> > > there
> > > > > have not been any changes to that website or any new
> releases of
> > > SurfStat.
> > > > >
> > > > > Tim
> > > > >
> > > > > > On May 23, 2019 at 4:36 PM Bruce Fischl <
> fis...@nmr.mgh.harvard.edu>
> > > > > wrote:
> > > > > >
> > > > > >
> > > > > > Hi Loic
> > > > > >
> > > > > > sorry, can you ask whoever maintains surfstat? Or
> maybe someone else
> > > on
> > > > > the
> > > > > > list knows? I don't. You can label surfaces manually
> with the dev
> > > > > version
> > > > > > of freeview if you want.
> > > > > >
> > > > > > cheers
> > > > > > Bruce
> > > > > >
> > > > > >
> > > > > > On Thu, 23 May 2019, Loïc
> > > > > > Daumail wrote:
> > > > > >
> > > > > > >
> > > > > > > External Email - Use Caution
> > > > > > >
> > > > > > > Hi Bruce,
> > > > > > >
> > > > > > > I am coming back to this request as I was
> wondering:
> > > > > > > would it be possible to load freesurfer surface
> atlases in
> > surfstat
> > > in
> > > > > order to label my surfaces?
> > > > > > > Is their format compatible? I saw surfstat only
> accepts a limited
> > > > > array of formats...
> > > > > > >
> > > > > > > Thanks,
> > > > > > > Best,
> > > > > > >
> > > > > > > Loïc
> > > > > > >
> > > > > > > Le lun. 20 mai 2019 à 16:15, Bruce Fischl
> > > <fis...@nmr.mgh.harvard.edu>
> > > > > a écrit :
> > > > > > > sure. All the fsaverages are in the same
> space (they are
> > > subsets
> > > > > of
> > > > > > > fsaverage)
> > > > > > > On Mon, 20 May 2019, Loïc Daumail wrote:
> > > > > > >
> > > > > > > >
> > > > > > > > External Email - Use Caution
> > > > > > > >
> > > > > > > > Hi Bruce,
> > > > > > > >
> > > > > > > > Thanks for your response. Is it also fine
> for fsaverage5
> > > > > surface or I need to switch
> > > > > > > my data to
> > > > > > > > fsaverage?
> > > > > > > >
> > > > > > > > Thanks,
> > > > > > > >
> > > > > > > > Sincerely,
> > > > > > > >
> > > > > > > > Loïc
> > > > > > > >
> > > > > > > > Le lun. 20 mai 2019 à 15:52, Bruce Fischl <
> > > > > fis...@nmr.mgh.harvard.edu> a écrit :
> > > > > > > > Hi Loic
> > > > > > > >
> > > > > > > > if you are on the fsaverage surface
> you can save the
> > > > > -log10(p) values
> > > > > > > > using save_mgh.m, then load them in
> freeview and
> > > > > threshold them or
> > > > > > > > whatever
> > > > > > > >
> > > > > > > > cheers
> > > > > > > > Bruce
> > > > > > > > On Mon, 20 May 2019, Loïc Daumail
> wrote:
> > > > > > > >
> > > > > > > > >
> > > > > > > > > External Email - Use
> Caution
> > > > > > > > >
> > > > > > > > > Also, one more thing:
> > > > > > > > >
> > > > > > > > > My data is on fsaverage5
> resolution.
> > > > > > > > >
> > > > > > > > > Thanks in advance,
> > > > > > > > >
> > > > > > > > > Sincerely,
> > > > > > > > >
> > > > > > > > > Loïc
> > > > > > > > >
> > > > > > > > > Le lun. 20 mai 2019 à 09:49, Loïc
> Daumail <
> > > > > loic.daum...@gmail.com> a écrit :
> > > > > > > > > Dear Freesurfer community,
> > > > > > > > >
> > > > > > > > > I obtained score matrices of
> surfaces after
> > > performing
> > > > > diffusion embedding map
> > > > > > > on fMRI
> > > > > > > > data.
> > > > > > > > >
> > > > > > > > > I am now using the SurfStat
> toolbox(Matlab) in
> > order
> > > > > to obtain p-value maps to
> > > > > > > compare
> > > > > > > > between
> > > > > > > > > two groups.
> > > > > > > > >
> > > > > > > > > After obtaining the p-value map, I
> would like to
> > > label
> > > > > the surface in order to
> > > > > > > > identify the
> > > > > > > > > regions of significant difference
> between both
> > > groups.
> > > > > > > > >
> > > > > > > > > Would it be possible to use
> Freview in order to do
> > > it?
> > > > > > > > > If so, which format should I use
> for my SurfStatP
> > > > > output p-value map?
> > > > > > > > > How would you do it, (tools to
> convert into a
> > > specific
> > > > > format, surfaces and
> > > > > > > atlas to
> > > > > > > > load on
> > > > > > > > > Freeview, which format for those
> ones)?
> > > > > > > > > I would be very grateful if you
> could detail a
> > bit,
> > > as
> > > > > I am new to these
> > > > > > > tools!
> > > > > > > > >
> > > > > > > > > Thanks in advance,
> > > > > > > > >
> > > > > > > > > Loïc Daumail
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> >_______________________________________________
> > > > > > > > Freesurfer mailing list
> > > > > > > > Freesurfer@nmr.mgh.harvard.edu
> > > > > > > >
> > > > >
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> > > > > > > >
> > > > > > > >
> > > > > > >
> >_______________________________________________
> > > > > > > Freesurfer mailing list
> > > > > > > Freesurfer@nmr.mgh.harvard.edu
> > > > > > >
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> > > > > > >
> > > > > > >
> > > > > > >_______________________________________________
> > > > > > Freesurfer mailing list
> > > > > > Freesurfer@nmr.mgh.harvard.edu
> > > > > >
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> > > > >
> > > > > --
> > > > > Dr. Tim Schäfer
> > > > > Postdoc Computational Neuroimaging
> > > > > Department of Child and Adolescent Psychiatry,
> Psychosomatics and
> > > > > Psychotherapy
> > > > > University Hospital Frankfurt, Goethe University
> Frankfurt am Main,
> > > Germany
> > > > >
> > > > > _______________________________________________
> > > > > Freesurfer mailing list
> > > > > Freesurfer@nmr.mgh.harvard.edu
> > > > >
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> > > > _______________________________________________
> > > > Freesurfer mailing list
> > > > Freesurfer@nmr.mgh.harvard.edu
> > > >
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> > >
> > > _______________________________________________
> > > Freesurfer mailing list
> > > Freesurfer@nmr.mgh.harvard.edu
> > >
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> > >
> > >
> > >_______________________________________________
> > Freesurfer mailing list
> > Freesurfer@nmr.mgh.harvard.edu
> > https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
> >
> >
> >_______________________________________________
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
