[Freesurfer] Post-doctoral fellowship: cortical integration timescales across the ventral stream
*Post-doctoral fellowship: cortical integration timescales across the ventral stream* We are looking for postdoc candidates for a European Union financed project investigating how temporal integration properties change at different levels of the processing hierarchy for meaningful objects, scenes and events. The fellow would work together with David Melcher and Scott Fairhall. Specifically, we are seeking a postdoc with advanced expertise in fMRI as demonstrated by two first author fMRI publications. Salary would be in the range of circa 24,000 – 30,000 euro (net) per year, commensurate with experience. The Center for Mind/Brain Sciences (CIMeC, http://web.unitn.it/en/cimec) at the University of Trento offers a vibrant research setting with state-of-the-art neuroimaging methodologies, including a research-only MRI scanner, MEG, EEG and TMS, as well as behavioral, eye tracking and motion tracking laboratories. Melcher Active Perception Lab: http://r.unitn.it/en/cimec/map FairLab: theFairLab.org http://thefairlab.org/ For informal inquiries about this position email: david (dot) melcher (AT) unitn (dot) it ___ Freesurfer mailing list Freesurfer@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.
Re: [Freesurfer] functional connectivity maps (correlation and z maps)
Thanks! Would you please explain why you think ces is the better map (vs z or
correlation coefficient)?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Friday, March 27, 2015 12:28 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z maps)
I think that the ces is the better map to use, but there's not a
consensus. For some reason that I don't understand, people starting
using the z-map or the correlation coefficient as the input to the group
level in contradiction to everything that had ever been done in
task-based analysis. In tests that I have done with task-based analysis,
I did not see much difference in using ces, cespct, z, pcc, or t.
doug
On 03/26/2015 09:21 AM, Barbour, Tracy,M.D. wrote:
Thanks Doug,
Hi,
I am wondering if you would give some input on this:
When we run the buckner scripts for functional connectivity analysis, those
scripts:
1. Obtain correlation maps of individual subjects, and convert them to
individual subject z-maps using Fisher's r-to-z transformation.
2. Obtain group-averaged correlation z-maps ( averaged across all subjects
in the group )
3. Apply an inverse Fisher's r-to-z transformation to the group-averaged
correlation z-map, yielding a group-averaged correlation map.
Using the fsFAST scripts I can:
1. Obtain the individual z maps (but these are converted from the sig.mgh map
not a correlation map).
2. Then I can concat all the individual z maps using isxconcat-sess by adding
the -map z flag. I get a group z.nii.gz map.
3. Now I am uncertain what to do for the within group analysis. The
mri_glmfit command can use the z.nii.gz file and the output is the sig.mgh
file. This is not analogous to the buckner scripts, as I am not converting
back to a correlation map, but am I getting the same result?
When I compare a within group analysis I get similar results if I use the
z.nii.gz or the ces.nii.gz as the input.
mri_glmfit --y ces.nii.gz --osgm --glmdir lh.wls.ces.ogsm --wls cesvar.nii.gz
--cortex --surf fsaverage lh --mask ../mask.nii.gz
When I compare the maps obtained during a regression analysis using the
z.nii.gz file and the ces.nii.gz file I get different results, so I'm not
sure which one to use.
mri_glmfit --y z.nii.gz --fsgd TAQ_ctl_blDIPS2008anatseed.fsgd --surf
fsaverage lh --glmdir lh.TAQ_ctl.glmdir --C ../slope.mtx --C ../intercept.mtx
--cortex
Which map should I use?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Wednesday, March 25, 2015 3:37 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z
maps)
cd glmdir/contrast
(could be sig.nii or sig.nii.gz)
matlab
sig = MRIread('sig.mgh');
p = (10.^-abs(sig.vol))/2;
z = sig;
z.vol = sign(sig.vol).*fast_p2z(p);
MRIwrite(z,'z.mgh')
On 03/25/2015 10:26 AM, Barbour, Tracy,M.D. wrote:
Hello,
I am doing a functional connectivity analysis. I want to make
individual pearson correlation maps for each subject (correlation
between the seed region and all other voxels in the brain) and then
convert them to a map of z-scores using a Fisher's z transformation. I
want to then get an averaged group z map.
I know there is an -m z option for isxconcat. From where are the z
values calculated from (Are the z values calculated from the ces map
or the pcc map)? How do I get the group z map?
Thank you in advance
Tracy
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
gr...@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/
___
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
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
gr...@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:
[Freesurfer] postdoc position in cognitive neuroscience
A Postdoctoral Research Fellow position is available at the Max Planck Institute for Human Cognitive and Brain Sciences (MPI-CBS) in Leipzig, Germany. The objective of the postdoctoral research project is to use high-resolution functional and structural magnetic resonance imaging (MRI) to understand the role of subcortical sensory structures in human perception and communication. The MPI-CBS is an internationally leading centre for cognitive and imaging neuroscience equipped with a 7.0 T MRI scanner, several 3.0 T MRI scanners, a 306 channels MEG system, TMS, tDCS, several EEG suites, and eye-tracking labs. All facilities and data analyses are supported by experienced IT specialists and physicists. Besides an excellent infrastructure, our institute offers an international and friendly environment with researchers from diverse backgrounds. The postdoc will be member of the group Neural Mechanisms of Human Communication led by Katharina von Kriegstein. The candidates must have a PhD (or equivalent) in neuroscience, experimental psychology, biology, or a related field, and should be able to demonstrate a consistently outstanding academic record, including publications. The ideal candidate will have expertise in the acquisition and analysis of neuroscientific data. Prior experience with high-resolution functional or structural MRI is preferred. The starting date for this position is flexible. Initially for two years, the position offers the possibility of extension for up to four years. Salary depends on experience and is based on regulations of the Max Planck Society. To apply, please include all documents in one PDF-file in the following order: CV, contact information for two references, a brief statement describing your personal qualifications and future research interests, copies of up to three of your publications. Applications with the subject heading HC15PD should be sent via email to: perso...@cbs.mpg.de. The deadline for application submission is 17 April 2015. Contact for informal enquiries regarding the post: Prof. Dr. Katharina von Kriegstein (kriegst...@cbs.mpg.de). For more information about the group see: http://www.cbs.mpg.de/groups/misc/humcomm. The MPI-CBS is an equal opportunities employer, committed to the advancement of individuals without regard to ethnicity, religion, gender, or disability. --- Katharina von Kriegstein Max Planck Research Group Leader Max Planck Institute for Human Cognitive and Brain Sciences Stephanstr. 1A, 04103 Leipzig, Germany Professor of Cognitive and Clinical Neuroscience Humboldt University of Berlin Rudower Chaussee 18, 12489 Berlin, Germany Phone +49 (0) 341-9940-2476 Fax +49 (0) 341-9940-2448 http://www.cbs.mpg.de/groups/misc/humcomm ___ Freesurfer mailing list Freesurfer@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.
Re: [Freesurfer] functional connectivity maps (correlation and z maps)
I think that the ces is the better map to use, but there's not a
consensus. For some reason that I don't understand, people starting
using the z-map or the correlation coefficient as the input to the group
level in contradiction to everything that had ever been done in
task-based analysis. In tests that I have done with task-based analysis,
I did not see much difference in using ces, cespct, z, pcc, or t.
doug
On 03/26/2015 09:21 AM, Barbour, Tracy,M.D. wrote:
Thanks Doug,
Hi,
I am wondering if you would give some input on this:
When we run the buckner scripts for functional connectivity analysis, those
scripts:
1. Obtain correlation maps of individual subjects, and convert them to
individual subject z-maps using Fisher's r-to-z transformation.
2. Obtain group-averaged correlation z-maps ( averaged across all subjects
in the group )
3. Apply an inverse Fisher's r-to-z transformation to the group-averaged
correlation z-map, yielding a group-averaged correlation map.
Using the fsFAST scripts I can:
1. Obtain the individual z maps (but these are converted from the sig.mgh map
not a correlation map).
2. Then I can concat all the individual z maps using isxconcat-sess by adding
the -map z flag. I get a group z.nii.gz map.
3. Now I am uncertain what to do for the within group analysis. The
mri_glmfit command can use the z.nii.gz file and the output is the sig.mgh
file. This is not analogous to the buckner scripts, as I am not converting
back to a correlation map, but am I getting the same result?
When I compare a within group analysis I get similar results if I use the
z.nii.gz or the ces.nii.gz as the input.
mri_glmfit --y ces.nii.gz --osgm --glmdir lh.wls.ces.ogsm --wls cesvar.nii.gz
--cortex --surf fsaverage lh --mask ../mask.nii.gz
When I compare the maps obtained during a regression analysis using the
z.nii.gz file and the ces.nii.gz file I get different results, so I'm not
sure which one to use.
mri_glmfit --y z.nii.gz --fsgd TAQ_ctl_blDIPS2008anatseed.fsgd --surf
fsaverage lh --glmdir lh.TAQ_ctl.glmdir --C ../slope.mtx --C ../intercept.mtx
--cortex
Which map should I use?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Wednesday, March 25, 2015 3:37 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z
maps)
cd glmdir/contrast
(could be sig.nii or sig.nii.gz)
matlab
sig = MRIread('sig.mgh');
p = (10.^-abs(sig.vol))/2;
z = sig;
z.vol = sign(sig.vol).*fast_p2z(p);
MRIwrite(z,'z.mgh')
On 03/25/2015 10:26 AM, Barbour, Tracy,M.D. wrote:
Hello,
I am doing a functional connectivity analysis. I want to make
individual pearson correlation maps for each subject (correlation
between the seed region and all other voxels in the brain) and then
convert them to a map of z-scores using a Fisher's z transformation. I
want to then get an averaged group z map.
I know there is an -m z option for isxconcat. From where are the z
values calculated from (Are the z values calculated from the ces map
or the pcc map)? How do I get the group z map?
Thank you in advance
Tracy
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
gr...@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/
___
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
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
gr...@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/
___
Freesurfer mailing list
Freesurfer@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
Re: [Freesurfer] error while loading shared libraries: libnewimage.so
It's solved, The problem has been solved by adding a line to .bashrc file. ... export FSL_DIR=/usr/share/fsl/5.0 ... Previously I manually corrected the FSL_DIR in the tcsh shell, were I run the Freesurfer, using: ... HP-Z620-RLavrador:~ setenv FSL_DIR /usr/share/fsl/5.0/ HP-Z620-RLavrador:~ source $FREESURFER_HOME/SetUpFreeSurfer.csh freesurfer-Linux-centos4_x86_64-stable-pub-v5.3.0 Setting up environment for FreeSurfer/FS-FAST (and FSL) FREESURFER_HOME /usr/local/freesurfer FSFAST_HOME /usr/local/freesurfer/fsfast FSF_OUTPUT_FORMAT nii.gz SUBJECTS_DIR /home/rlavrador/proj/Freesurfer MNI_DIR /usr/local/freesurfer/mni FSL_DIR /usr/share/fsl/5.0/ ... but it did not solve the problem. I had to add the line to .bashrc in order to solve the problem, as I explained. It worked for me! Best regards, 2015-03-27 14:10 GMT+00:00 Rui Lavrador ruilavra...@gmail.com: Hello, I'm trying to prepare my diffusion data with track-all -prep but trac-preproc exited with ERRORS I think it is related to the FSL instructions, because it gives this message in the beginning: ... flirt: error while loading shared libraries: libnewimage.so: cannot open shared object file: No such file or directory trac-preproc -c /home/rlavrador/proj/Freesurfer_diff/AAA/scripts/dmrirc.local -log /home/rlavrador/proj/Freesurfer_diff/AAA/scripts/trac-all.log -cmd /home/rlavrador/proj/Freesurfer_diff/AAA/scripts/trac-all.cmd ... and it also cannot run the flip4fsl ... flip4fsl /home/rlavrador/proj/Freesurfer_diff/AAA/dmri/dwi_orig.nii.gz /home/rlavrador/proj/Freesurfer_diff/AAA/dmri/dwi_orig_flip.nii.gz INFO: input image orientation is LPS INFO: input image determinant is 8 fslswapdim /home/rlavrador/proj/Freesurfer_diff/AAA/dmri/dwi_orig.nii.gz x -y z /home/rlavrador/proj/Freesurfer_diff/AAA/dmri/dwi_orig_flip.nii.gz fslorient -forceradiological /home/rlavrador/proj/Freesurfer_diff/AAA/dmri/dwi_orig_flip.nii.gz fslorient: error while loading shared libraries: libnewimage.so: cannot open shared object file: No such file or directory ... and consequently the eddy correct: ... eddy_correct /home/rlavrador/proj/Freesurfer_diff/MCD20111024/dmri/dwi_orig_flip.nii.gz /home/rlavrador/proj/Freesurfer_diff/MCD20111024/dmri/dwi.nii.gz 0 Input does not exist or is not in a supported format ... I think I have the variables well configured ... HP-Z620-RLavrador:~ source $FREESURFER_HOME/SetUpFreeSurfer.csh freesurfer-Linux-centos4_x86_64-stable-pub-v5.3.0 Setting up environment for FreeSurfer/FS-FAST (and FSL) FREESURFER_HOME /usr/local/freesurfer FSFAST_HOME /usr/local/freesurfer/fsfast FSF_OUTPUT_FORMAT nii.gz SUBJECTS_DIR /home/rlavrador/proj/Freesurfer MNI_DIR /usr/local/freesurfer/mni FSL_DIR /usr/share/fsl/5.0/ ... I would like to know what I am doing wrong. Some clues? Regards, -- Rui Lavrador Visual Neuroscience Laboratory IBILI-Faculdade de Medicina Azinhaga de Santa Comba 3000-548 Coimbra, Portugal Phone: +351 964985485 e-mail: rui.lavra...@fmed.uc.pt e-mail2: ruilavra...@gmail.com -- Rui Lavrador Visual Neuroscience Laboratory IBILI-Faculdade de Medicina Azinhaga de Santa Comba 3000-548 Coimbra, Portugal Phone: +351 964985485 e-mail: rui.lavra...@fmed.uc.pt e-mail2: ruilavra...@gmail.com ___ Freesurfer mailing list Freesurfer@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.
Re: [Freesurfer] Tracula: missing or incomplete tracts
Ops... I did not see these replies before! Thank you, both!! Christopher, the command line works great. I also found the --viewport and the --ss, which is exactly what I was looking for! I will try the dev version that Anastasia suggested. Thanks a lot, Amelia -Original Message- From: freesurfer-boun...@nmr.mgh.harvard.edu [mailto:freesurfer-boun...@nmr.mgh.harvard.edu] On Behalf Of Watson, Christopher Sent: Tuesday, March 17, 2015 9:33 PM To: Freesurfer support list; Ruopeng Wang Subject: Re: [Freesurfer] Tracula: missing or incomplete tracts What about: freeview -v FA.nii.gz -v V1.nii.gz:vector=yes From: freesurfer-boun...@nmr.mgh.harvard.edu [freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Anastasia Yendiki [ayend...@nmr.mgh.harvard.edu] Sent: Tuesday, March 17, 2015 11:57 AM To: Freesurfer support list; Ruopeng Wang Subject: Re: [Freesurfer] Tracula: missing or incomplete tracts Hi Amelia - Sorry, not that I know of. Perhaps this can be a feature request for Ruopeng to add to freeview :) BTW, freeview in 5.3 had a bug with the vector display for some input image orientations, but this has been fixed in the dev version (that you can download on the web site). a.y On Thu, 12 Mar 2015, Versace, Amelia wrote: Dear Anastasia, I was wondering if there is a way to automatically (in command line) derive the FA/V1 vector image. fslview FA.nii.gz V1.nii.gz allows to load the FA and V1 image, but then manual section of the (lines) is needed. freeview -dti V1 FA does it too (with different xyz convention), but the 'display as vectors' needs to be manually checked. Is there any way to produce this image using command line only? Thanks a lot!! Amelia -Original Message- From: freesurfer-boun...@nmr.mgh.harvard.edu [mailto:freesurfer-boun...@nmr.mgh.harvard.edu] On Behalf Of Anastasia Yendiki Sent: Friday, March 06, 2015 6:56 PM To: Freesurfer support list Subject: Re: [Freesurfer] Tracula: missing or incomplete tracts Hi Eileen - It looks like there was a A-P flip introduced to your gradient vectors. (See screenshot of the tensor eigenvectors. Because they look right in the coronal view but wrong in the sagittal and axial views, that's why I'm assuming the flip is in the A-P direction.) To fix this you need to multiply the y coordinate of your gradient vectors with -1 and rerun. If there are still any cases of missing tracts, please look at the bottom of your dmrirc file for instructions on how to use the reinit parameter (or search for previous emails on that parameter in the archives). Note that your DWI voxel size is anisotropic (finer resolution in x/y than z), this may cause bias in tractography, so I'd recommend that you switch to isotropic resolution for future acquisitions if you have control over it. Hope this helps, a.y On Fri, 27 Feb 2015, Eileen Moore wrote: Hi Anastasia, Sorry for the delay -- I missed your response. I did look at the anatomical segmentations and thought they looked OK. I've uploaded the tracula and corresponding freesurfer data for two subjects: one with complete tracts and one with missing/incomplete tracts. I also uploaded my dmrirc file. Thanks for taking a look for me. Eileen. Hi Eileen - Have you checked the anatomical segmentations of the subjects? The mask that's used by default is the aparc+aseg_mask, which comes from registering the cortical parcellationg and subcortical segmentation from T1 to diffusion space, and then dilating it by a couple of voxels. If you upload an example data set for me here, I can take a look: https://urldefense.proofpoint.com/v2/url?u=https-3A__gate.nmr.mgh.har vard.edu_filedrop2_d=BQIDaQc=qS4goWBT7poplM69zy_3xhKwEW14JZMSdioCop pxeFUr=bw42CATdUgaQuKlfBniyB_kCJBdQgH3uC9uSwd5UzJ12A5jbDc-vHb9P3-hkq 5C4m=T2p7z8F4umjzf0YkkzlBSH_xnwrFLJ1rKFdoTAMUxbMs=TVmCtWOX7TmzlQhiV JibKKVtj6CQczyAkyupGnmGZa0e= Please include all tracula-related directories of the subject (dmri, dlabel, etc). Thanks! a.y On Thu, 18 Dec 2014, Eileen Moore wrote: Hi - I'm having difficulty with missing or incomplete tracts most of my subjects. I'm hoping for suggestions on where I can look for data problems. The majority of my subjects have at least one missing/incomplete tract, but the specific problematic tract varies across subjects (e.g., one subject has a missing L.Uncinate; another subject has a missing Forceps Major; another is missing the ILF bilaterally, etc.). For my problematic tracts, the path.pd.nii.gz is a single line/curve rather than the diffuse volumetric distribution. I'm not sure how to correct this. I have checked my eigenvectors -- the lines appear to be pointing in the correct directions in my dtifit_V1 -- so I believe my gradient table is correct I've checked my images for motion via visual inspection and by excluding any subjects with dwi_motion
Re: [Freesurfer] functional connectivity maps (correlation and z maps)
Thanks! Are both the ces and cespct normalized? Is there a reason you would use
one over the other?
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Friday, March 27, 2015 3:15 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z maps)
The statistic maps (z, t, correlation coefficient) depend on the noise
of the individual analysis, ie, two subjects could have identical
correlation but have different correlation coefficient or z because one
is noisier than the other (eg, moves more, physio is different). This
means that you can find difference between groups because one group is
noisier than the other. The ces and cespct do not have this problem (or,
rather, it is much reduced).
doug
On 03/27/2015 01:53 PM, Barbour, Tracy,M.D. wrote:
Thanks! Would you please explain why you think ces is the better map (vs z or
correlation coefficient)?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Friday, March 27, 2015 12:28 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z
maps)
I think that the ces is the better map to use, but there's not a
consensus. For some reason that I don't understand, people starting
using the z-map or the correlation coefficient as the input to the group
level in contradiction to everything that had ever been done in
task-based analysis. In tests that I have done with task-based analysis,
I did not see much difference in using ces, cespct, z, pcc, or t.
doug
On 03/26/2015 09:21 AM, Barbour, Tracy,M.D. wrote:
Thanks Doug,
Hi,
I am wondering if you would give some input on this:
When we run the buckner scripts for functional connectivity analysis, those
scripts:
1. Obtain correlation maps of individual subjects, and convert them to
individual subject z-maps using Fisher's r-to-z transformation.
2. Obtain group-averaged correlation z-maps ( averaged across all subjects
in the group )
3. Apply an inverse Fisher's r-to-z transformation to the group-averaged
correlation z-map, yielding a group-averaged correlation map.
Using the fsFAST scripts I can:
1. Obtain the individual z maps (but these are converted from the sig.mgh
map not a correlation map).
2. Then I can concat all the individual z maps using isxconcat-sess by
adding the -map z flag. I get a group z.nii.gz map.
3. Now I am uncertain what to do for the within group analysis. The
mri_glmfit command can use the z.nii.gz file and the output is the sig.mgh
file. This is not analogous to the buckner scripts, as I am not converting
back to a correlation map, but am I getting the same result?
When I compare a within group analysis I get similar results if I use the
z.nii.gz or the ces.nii.gz as the input.
mri_glmfit --y ces.nii.gz --osgm --glmdir lh.wls.ces.ogsm --wls
cesvar.nii.gz --cortex --surf fsaverage lh --mask ../mask.nii.gz
When I compare the maps obtained during a regression analysis using the
z.nii.gz file and the ces.nii.gz file I get different results, so I'm not
sure which one to use.
mri_glmfit --y z.nii.gz --fsgd TAQ_ctl_blDIPS2008anatseed.fsgd --surf
fsaverage lh --glmdir lh.TAQ_ctl.glmdir --C ../slope.mtx --C
../intercept.mtx --cortex
Which map should I use?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Wednesday, March 25, 2015 3:37 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z
maps)
cd glmdir/contrast
(could be sig.nii or sig.nii.gz)
matlab
sig = MRIread('sig.mgh');
p = (10.^-abs(sig.vol))/2;
z = sig;
z.vol = sign(sig.vol).*fast_p2z(p);
MRIwrite(z,'z.mgh')
On 03/25/2015 10:26 AM, Barbour, Tracy,M.D. wrote:
Hello,
I am doing a functional connectivity analysis. I want to make
individual pearson correlation maps for each subject (correlation
between the seed region and all other voxels in the brain) and then
convert them to a map of z-scores using a Fisher's z transformation. I
want to then get an averaged group z map.
I know there is an -m z option for isxconcat. From where are the z
values calculated from (Are the z values calculated from the ces map
or the pcc map)? How do I get the group z map?
Thank you in advance
Tracy
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
Re: [Freesurfer] mris_preproc output file PET data
Hi Alex, mris_preproc produces a stack images (27 in the first case and 16 in the 2nd case). For some reason the first two images are the same in both stacks. This is the way that you have run the program because both calls to mris_preproc use NNC0035_pet_lh.mgh and NNC0045_pet_lh.mgh as the first two inputs doug On 03/27/2015 02:32 PM, Alexandra Tanner wrote: Hi Doug and Freesurfers, Our group has PET data for 27 subjects that we are hoping to combine to generate surface group average maps. We would also like to generate average maps for 3 subsets of subjects based on genotype. I ran mris_preproc to combine my individual subject PET surface maps to generate a concatenated map for my N=27 subjects and 3 separate subgroups based on genotype (N=16, N=9, N=2), with the commands below. We viewed the concatenated maps generated by mris_preproc in tksurfer and the N=27 and N=16 maps are identical, which we find puzzling (please see images attached). We are curious 1) what the output map of the mris_preproc command actually is (ex. a sum, an average, etc), and 2) if anyone knows why our outputs would be identical for our 27 subjects and our 16 subjects. Any help or explanation would be greatly appreciated! Best, Alex mris_preproc commands run: setenv SUBJECTS_DIR /cluster/roffman/users/Stable5_PerRun cd /autofs/cluster/roffman/users/Stable5_PerRun/NNC_PMOD_LoganRef_BPnd/PET_analyses /usr/local/freesurfer/stable5_0_0/bin/mris_preproc --target fsaverage --hemi lh --is NNC0035_pet_lh.mgh --is NNC0045_pet_lh.mgh --is NNC0047_pet_lh.mgh --is NNC0065_pet_lh.mgh --is NNC0078_pet_lh.mgh --is NNC0089_pet_lh.mgh --is NNC0095_pet_lh.mgh --is NNC0101_pet_lh.mgh --is NNC0102_pet_lh.mgh --is NNC0114_pet_lh.mgh --is NNC0118_pet_lh.mgh --is NNC0119_pet_lh.mgh --is NNC0901_pet_lh.mgh --is NNC0902_pet_lh.mgh --is NNC0906_pet_lh.mgh --is NNC0908_pet_lh.mgh --is NNC0915_pet_lh.mgh --is NNC0916_pet_lh.mgh --is NNC0917_pet_lh.mgh --is NNC0918_pet_lh.mgh --is NNC0919_pet_lh.mgh --is NNC0924_pet_lh.mgh --is NNC0926_pet_lh.mgh --is NNC0928_pet_lh.mgh --is NNC0930_pet_lh.mgh --is NNC0931_pet_lh.mgh --is NNC0932_pet_lh.mgh --f /cluster/roffman/users/Stable5_PerRun/Subject_Files/27_NNC_Subjects --out /cluster/roffman/users/Stable5_PerRun/NNC_PMOD_LoganRef_BPnd/PET_analyses/27_NNC_removed_bad_PET_fMRI/lh.pet.mgh setenv SUBJECTS_DIR /cluster/roffman/users/Stable5_PerRun/ cd /autofs/cluster/roffman/users/Stable5_PerRun/NNC_PMOD_LoganRef_BPnd/PET_analyses /usr/local/freesurfer/stable5_0_0/bin/mris_preproc --target fsaverage --hemi lh --is NNC0035_pet_lh.mgh --is NNC0045_pet_lh.mgh --is NNC0065_pet_lh.mgh --is NNC0089_pet_lh.mgh --is NNC0118_pet_lh.mgh --is NNC0901_pet_lh.mgh --is NNC0902_pet_lh.mgh --is NNC0908_pet_lh.mgh --is NNC0915_pet_lh.mgh --is NNC0916_pet_lh.mgh --is NNC0917_pet_lh.mgh --is NNC0924_pet_lh.mgh --is NNC0928_pet_lh.mgh --is NNC0930_pet_lh.mgh --is NNC0931_pet_lh.mgh --is NNC0932_pet_lh.mgh --f /cluster/roffman/users/Stable5_PerRun/Subject_Files/16_NNC_AGCarrier_COMT_27NNC --out /cluster/roffman/users/Stable5_PerRun/NNC_PMOD_LoganRef_BPnd/PET_analyses/16_AG_Carrier_COMT_27NNC/lh.pet.mgh -- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@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/ ___ Freesurfer mailing list Freesurfer@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.
Re: [Freesurfer] -per-run analysis in native space
Yes. Note that that will produce huge files (float images 256^3 by number of time points). doug On 03/27/2015 01:50 PM, SHAHIN NASR wrote: Hi Surfers, I want to find a way to analyze subcortical activities in native space and still take advantage of per-run registration. As far as I understand, this is not possible in current preproc-sess. We need one extra step in which we have to map all fmcpr.nii.gz files to subject's orig.mgz. I am trying: mri_vol2vol --mov fmcpr.nii.gz --o fmcpr.native.nii.gz --fstarg orig.mgz --reg register.dof6.dat --no-save-reg --interp trilin in my mind, it is equivalent of what preproc-sess does when we use -mni305 flag (it uses mri_vol2vol to map fmcpr to mni space). Am I right? -- Shahin Nasr PhD in Cognitive Neuroscience Martinos Imaging Center, MGH Harvard Medical School -- Douglas N. Greve, Ph.D. MGH-NMR Center gr...@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/ ___ Freesurfer mailing list Freesurfer@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.
Re: [Freesurfer] Using a2009s.annot as seed masks for tractography
On Thu, 26 Mar 2015, André Schmidt wrote: Hi Andre, Try using the --seg flag with vol2label, as in: mri_label2vol --seg left_putamen2.nii.gz --temp b0_brain.nii.gz --reg registration2b0.dat --o left_putamen_dwi_seed left_putamen2.nii.gz is a segmentation, where each voxel in the volume is assigned a number (in your case 1 for left putamen and 0 everywhere else). Freesurfer label files have a different format. Run mri_label2vol --help for more information. Hope this helps, and let us know if there are more problems. Best, Lee Hi Anastasia, Sorry for not being clearer. According to your email, I have labelled all my regions of interest using this command: mri_binarize --match ID --i aparc.a2009s+aseg.mgz --o labelID.nii.gz Let's take the left putamen as an example: cmdline mri_binarize --match 12 --i aparc+aseg.mgz --o left_putamen2.nii.gz sysname Darwin hostname Andres-MacBook-Pro.local machine x86_64 user andre input aparc+aseg.mgz frame 0 nErode3d 0 nErode2d 0 output left_putamen2.nii.gz Binarizing based on matching values nMatch 1 012 binval1 binvalnot 0 Found 6228 values in range Counting number of voxels Found 6228 voxels in final mask mri_binarize done Please find attached the output file (left_putamen2.nii.gz). After this, I used the following command to create the register.dat file: bbregister --s HC001 --mov HC001/mri/bo_brain.nii.gz --reg HC001/mri/registration2b0.dat --dti --init-fsl Please find attached the registration2b0.dat file. Then I wanted to convert the label file into diffusion space with mri_label2vol with this command: mri_label2vol --label left_putamen2.nii.gz --temp b0_brain.nii.gz --reg registration2b0.dat --o left_putamen_dwi_seed Number of labels: 1 left_putamen2.nii.gz Annot File: (null) Template Volume: b0_brain.nii.gz Outut Volume: left_putamen_dwi_seed Registration File: registration2b0.dat Fill Threshold: 0 Label Vox Vol: 1 ProjType: (null) ProjTypeId: 0 ProjStart: 0 ProjStop: 0 ProjDelta: 0.1 Subject: (null) Hemi: (null) UseNewASeg2Vol: 1 DoLabelStatVol 0 LabelCodeOffset 0 setenv SUBJECTS_DIR /Applications/freesurfer/subjects $Id: mri_label2vol.c,v 1.34.2.5 2012/06/08 17:31:03 greve Exp $ Template RAS-to-Vox: -0.400 0.000 0.000 64.000; -0.000 -0.000 -0.400 48.000; -0.000 0.400 -0.000 27.000; 0.000 0.000 0.000 1.000; Template Voxel Volume: 15.625 nHits Thresh: 0 Loading registration from registration2b0.dat RegMat: 0.996 0.088 -0.024 3.909; 0.025 -0.007 1.000 0.627; 0.088 -0.996 -0.009 -10.557; 0.000 0.000 0.000 1.000; Label RAS-to-Vox: -0.398 -0.035 0.010 62.436; -0.035 0.398 0.004 52.223; 0.010 -0.003 0.400 27.251; 0.000 0.000 0.000 1.000; nlabels = 1 Allocating Hit Volume (663552) voxels Loading left_putamen2.nii.gz ??? mri_label2vol: could not scan # of lines from label file ERROR reading left_putamen2.nii.gz Obviously something is wrong with the label file - do you see what I am doing wrong? Thank you Andre Von: freesurfer-boun...@nmr.mgh.harvard.edu [freesurfer-boun...@nmr.mgh.harvard.edu]quot; im Auftrag von quot;Anastasia Yendiki [ayend...@nmr.mgh.harvard.edu] Gesendet: Dienstag, 17. März 2015 19:43 An: Freesurfer support list Betreff: Re: [Freesurfer] Using a2009s.annot as seed masks for tractography Hi Andre - There is no way for us to guess what went wrong if we don't know the output of the command that you're trying to run. For example, something went wrong could mean that that the command failed without creating any volume, that the volume was created but it doesn't look right, etc. Please send us the command line that you type and all the text that the command outputs on your terminal window. a.y On Tue, 17 Mar 2015, André Schmidt wrote: Hi Anatasia, I was able to create a registration.dat with bbregister now
Re: [Freesurfer] functional connectivity maps (correlation and z maps)
The statistic maps (z, t, correlation coefficient) depend on the noise
of the individual analysis, ie, two subjects could have identical
correlation but have different correlation coefficient or z because one
is noisier than the other (eg, moves more, physio is different). This
means that you can find difference between groups because one group is
noisier than the other. The ces and cespct do not have this problem (or,
rather, it is much reduced).
doug
On 03/27/2015 01:53 PM, Barbour, Tracy,M.D. wrote:
Thanks! Would you please explain why you think ces is the better map (vs z or
correlation coefficient)?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Friday, March 27, 2015 12:28 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z
maps)
I think that the ces is the better map to use, but there's not a
consensus. For some reason that I don't understand, people starting
using the z-map or the correlation coefficient as the input to the group
level in contradiction to everything that had ever been done in
task-based analysis. In tests that I have done with task-based analysis,
I did not see much difference in using ces, cespct, z, pcc, or t.
doug
On 03/26/2015 09:21 AM, Barbour, Tracy,M.D. wrote:
Thanks Doug,
Hi,
I am wondering if you would give some input on this:
When we run the buckner scripts for functional connectivity analysis, those
scripts:
1. Obtain correlation maps of individual subjects, and convert them to
individual subject z-maps using Fisher's r-to-z transformation.
2. Obtain group-averaged correlation z-maps ( averaged across all subjects
in the group )
3. Apply an inverse Fisher's r-to-z transformation to the group-averaged
correlation z-map, yielding a group-averaged correlation map.
Using the fsFAST scripts I can:
1. Obtain the individual z maps (but these are converted from the sig.mgh
map not a correlation map).
2. Then I can concat all the individual z maps using isxconcat-sess by
adding the -map z flag. I get a group z.nii.gz map.
3. Now I am uncertain what to do for the within group analysis. The
mri_glmfit command can use the z.nii.gz file and the output is the sig.mgh
file. This is not analogous to the buckner scripts, as I am not converting
back to a correlation map, but am I getting the same result?
When I compare a within group analysis I get similar results if I use the
z.nii.gz or the ces.nii.gz as the input.
mri_glmfit --y ces.nii.gz --osgm --glmdir lh.wls.ces.ogsm --wls
cesvar.nii.gz --cortex --surf fsaverage lh --mask ../mask.nii.gz
When I compare the maps obtained during a regression analysis using the
z.nii.gz file and the ces.nii.gz file I get different results, so I'm not
sure which one to use.
mri_glmfit --y z.nii.gz --fsgd TAQ_ctl_blDIPS2008anatseed.fsgd --surf
fsaverage lh --glmdir lh.TAQ_ctl.glmdir --C ../slope.mtx --C
../intercept.mtx --cortex
Which map should I use?
Thank you
Tracy
From: freesurfer-boun...@nmr.mgh.harvard.edu
[freesurfer-boun...@nmr.mgh.harvard.edu] on behalf of Douglas N Greve
[gr...@nmr.mgh.harvard.edu]
Sent: Wednesday, March 25, 2015 3:37 PM
To: freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] functional connectivity maps (correlation and z
maps)
cd glmdir/contrast
(could be sig.nii or sig.nii.gz)
matlab
sig = MRIread('sig.mgh');
p = (10.^-abs(sig.vol))/2;
z = sig;
z.vol = sign(sig.vol).*fast_p2z(p);
MRIwrite(z,'z.mgh')
On 03/25/2015 10:26 AM, Barbour, Tracy,M.D. wrote:
Hello,
I am doing a functional connectivity analysis. I want to make
individual pearson correlation maps for each subject (correlation
between the seed region and all other voxels in the brain) and then
convert them to a map of z-scores using a Fisher's z transformation. I
want to then get an averaged group z map.
I know there is an -m z option for isxconcat. From where are the z
values calculated from (Are the z values calculated from the ces map
or the pcc map)? How do I get the group z map?
Thank you in advance
Tracy
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
--
Douglas N. Greve, Ph.D.
MGH-NMR Center
gr...@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/
___
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
