Hi Alex,

try "volume scanner coords" for the subjects that don't have the MNI coords. Sorry, I know this is more confusing than it needs to be. In the new version you just specify voxel coords :)

Bruce On Thu, 30 Mar 2006, Fornito, Alexander wrote:

Hi, I need to manually specify a seed point for the corpus callosum cutting plane on some of my images, although some of them don't seem to have the MNI coordiates available through clicking view >information in tkmedit. Not sure why this is the case, since other images do have this info. Specifying the RAS coord doens't fix the problem. I'm using recon-all version 1.104 2005/09/07 (I'm trying to finish off a study using the dev version and wold dearly like to avoid re-processing everyone with a newer version!!). Any help is much appreciated. Many thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Hi Bruce, I tried using the volume scanner coords (which were identical to the RAS coords), but this didn't seem to do the trick; the cutting plane was still skewed away from the true midline. After re-running some of my images using the volume scanner coords, I've noticed that I now have the MNI coords. However, I still didn't get them for some images. Is there a flag or command that I can specify to ensure that I do obtain the MNI coords? Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Thu 3/30/2006 11:41 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] MNI coords

Hi Alex,

try "volume scanner coords" for the subjects that don't have the MNI coords. Sorry, I know this is more confusing than it needs to be. In the new version you just specify voxel coords :)

Bruce On Thu, 30 Mar 2006, Fornito, Alexander wrote:

Hi, I need to manually specify a seed point for the corpus callosum cutting plane on some of my images, although some of them don't seem to have the MNI coordiates available through clicking view >information in tkmedit. Not sure why this is the case, since other images do have this info. Specifying the RAS coord doens't fix the problem. I'm using recon-all version 1.104 2005/09/07 (I'm trying to finish off a study using the dev version and wold dearly like to avoid re-processing everyone with a newer version!!). Any help is much appreciated. Many thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

I'll give it a try. What should the input and output vols be if I want to use the coords to place seed points? Does it matter which vol I use?

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

________________________________

From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Thu 4/6/2006 11:03 PM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] MNI coords

Hi Alex,

you should get the MNI coords for all your subjects - recon-all runs the MNI talairach procedure. Did you move things around on disk? I think there are times that it can't find the xform anymore because it is in a different location than when you ran recon-all. See if there is a talairach.xfm in the subjects mri/transforms directory for one of the ones you don't get MNI coords. If it's there and you're not getting MNI coords, try using

mri_add_xform_to_header <full path to talairach.xfm> <input vol> <output vol>

cheers, Bruce

On Thu, 6 Apr 2006, Fornito, Alexander wrote:

Hi Bruce, I tried using the volume scanner coords (which were identical to the RAS coords), but this didn't seem to do the trick; the cutting plane was still skewed away from the true midline. After re-running some of my images using the volume scanner coords, I've noticed that I now have the MNI coords. However, I still didn't get them for some images. Is there a flag or command that I can specify to ensure that I do obtain the MNI coords? Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Thu 3/30/2006 11:41 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] MNI coords

Hi Alex,

try "volume scanner coords" for the subjects that don't have the MNI coords. Sorry, I know this is more confusing than it needs to be. In the new version you just specify voxel coords :)

Bruce On Thu, 30 Mar 2006, Fornito, Alexander wrote:

...

Hi, I need to manually specify a seed point for the corpus callosum cutting plane on some of my images, although some of them don't seem to have the MNI coordiates available through clicking view >information in tkmedit. Not sure why this is the case, since other images do have this info. Specifying the RAS coord doens't fix the problem. I'm using recon-all version 1.104 2005/09/07 (I'm trying to finish off a study using the dev version and wold dearly like to avoid re-processing everyone with a newer version!!). Any help is much appreciated. Many thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

do it on the volume you want to view (or all of them). T1.mgz, orig.mgz, etc...

On Fri, 7 Apr 2006, Fornito, Alexander wrote:

I'll give it a try. What should the input and output vols be if I want to use the coords to place seed points? Does it matter which vol I use?

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Thu 4/6/2006 11:03 PM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] MNI coords

Hi Alex,

you should get the MNI coords for all your subjects - recon-all runs the MNI talairach procedure. Did you move things around on disk? I think there are times that it can't find the xform anymore because it is in a different location than when you ran recon-all. See if there is a talairach.xfm in the subjects mri/transforms directory for one of the ones you don't get MNI coords. If it's there and you're not getting MNI coords, try using

mri_add_xform_to_header <full path to talairach.xfm> <input vol> <output vol>

cheers, Bruce

On Thu, 6 Apr 2006, Fornito, Alexander wrote:

...

Hi Bruce, I tried using the volume scanner coords (which were identical to the RAS coords), but this didn't seem to do the trick; the cutting plane was still skewed away from the true midline. After re-running some of my images using the volume scanner coords, I've noticed that I now have the MNI coords. However, I still didn't get them for some images. Is there a flag or command that I can specify to ensure that I do obtain the MNI coords? Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: Thu 3/30/2006 11:41 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] MNI coords

Hi Alex,

try "volume scanner coords" for the subjects that don't have the MNI coords. Sorry, I know this is more confusing than it needs to be. In the new version you just specify voxel coords :)

Bruce On Thu, 30 Mar 2006, Fornito, Alexander wrote:

...

Hi, I need to manually specify a seed point for the corpus callosum cutting plane on some of my images, although some of them don't seem to have the MNI coordiates available through clicking view >information in tkmedit. Not sure why this is the case, since other images do have this info. Specifying the RAS coord doens't fix the problem. I'm using recon-all version 1.104 2005/09/07 (I'm trying to finish off a study using the dev version and wold dearly like to avoid re-processing everyone with a newer version!!). Any help is much appreciated. Many thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Hi, I'd like to run a simple repeated-measures analysis tetsing for change in thickness measures between 2 time points. There is no between-subjects factor, only the effect of time. Do I just set up a standard DOSS FSGD file, or do I use a different format? The tutorial only covers between-subjects designs. Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

Hi Doug, Is mri_preproc a new addition to the stable version? I'm using recon-all-nmr,v 1.104 2005/09/07 and can't seem to find it. I tried running the analysis through mris_glm as per option 1 as you described, but came across a problem. Command, error, and my FSGD are below. Thanks, Alex

Command:

mris_glm --surfmeas thickness --hemi lh --trgsubj ../average --fsgd ./test_fsgd.txt doss --beta ./beta_doss_thickness-100_lh bfloat --var ./var_doss_thickness-100_lh.w paint --y ./y_doss_thickness-100_lh_000.bfloat --nsmooth 100 gdfReadHeader: reading ./test_fsgd.txt INFO: ignoring tag 1 INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

My FSGD:

GroupDescriptorFile 1 Title Time 1 vs Time 2 Class class_time1_subj1 Class class_time1_subj2 Class class_time1_subj3 Class class_time1_subj4 Class class_time1_subj5 Class class_time2_subj1 Class class_time2_subj2 Class class_time2_subj3 Class class_time2_subj4 Class class_time2_subj5 Variables Time Input time1_subj1 class_time1_subj1 -1 Input time1_subj2 class_time1_subj2 -1 Input time1_subj3 class_time1_subj3 -1 Input time1_subj4 class_time1_subj4 -1 Input time1_subj5 class_time1_subj5 -1 Input time2_subj1 class_time2_subj1 1 Input time2_subj2 class_time2_subj2 1 Input time2_subj3 class_time2_subj3 1 Input time2_subj4 class_time2_subj4 1 Input time2_subj5 class_time2_subj5 1 DefaultVariable Time

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 9:31 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

There are two ways to do it:

1. You can use DOSS, making each subject it's own class. Then create one "Variable". For a given input, set the variable to either +1 or -1 depending upon whether it corresponds to the first or second time point. Then create a contrast as [0 0 0 0 ... 0 0 0 1], where there are as many 0s as subjects. The last "1" will correspond to the Time Point variable.

2. Alternatively, if you are using mris_preproc, you can specify "--paired-diff", then just use a one-sample group mean test. See the --help. This may be easier, and will also allow you to add a between subject factor.

doug

Fornito, Alexander wrote:

Hi, I'd like to run a simple repeated-measures analysis tetsing for change in thickness measures between 2 time points. There is no between-subjects factor, only the effect of time. Do I just set up a standard DOSS FSGD file, or do I use a different format? The tutorial only covers between-subjects designs. Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

I still get the same error message. Maybe I've coded the variable wrong?

gdfReadHeader: reading ./test_fsgd.txt INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 11:43 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] Repeated-measures in mris_glm

Yes, mris_preproc is new with stable. In your FSGD file, try specifying the title as a single string without spaces.

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

Hi Doug, Is mri_preproc a new addition to the stable version? I'm using recon-all-nmr,v 1.104 2005/09/07 and can't seem to find it. I tried running the analysis through mris_glm as per option 1 as you described, but came across a problem. Command, error, and my FSGD are below. Thanks, Alex

Command:

mris_glm --surfmeas thickness --hemi lh --trgsubj ../average --fsgd ./test_fsgd.txt doss --beta ./beta_doss_thickness-100_lh bfloat --var ./var_doss_thickness-100_lh.w paint --y ./y_doss_thickness-100_lh_000.bfloat --nsmooth 100 gdfReadHeader: reading ./test_fsgd.txt INFO: ignoring tag 1 INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

My FSGD:

GroupDescriptorFile 1 Title Time 1 vs Time 2 Class class_time1_subj1 Class class_time1_subj2 Class class_time1_subj3 Class class_time1_subj4 Class class_time1_subj5 Class class_time2_subj1 Class class_time2_subj2 Class class_time2_subj3 Class class_time2_subj4 Class class_time2_subj5 Variables Time Input time1_subj1 class_time1_subj1 -1 Input time1_subj2 class_time1_subj2 -1 Input time1_subj3 class_time1_subj3 -1 Input time1_subj4 class_time1_subj4 -1 Input time1_subj5 class_time1_subj5 -1 Input time2_subj1 class_time2_subj1 1 Input time2_subj2 class_time2_subj2 1 Input time2_subj3 class_time2_subj3 1 Input time2_subj4 class_time2_subj4 1 Input time2_subj5 class_time2_subj5 1 DefaultVariable Time

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 9:31 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

There are two ways to do it:

1. You can use DOSS, making each subject it's own class. Then create one

"Variable". For a given input, set the variable to either +1 or -1 depending upon whether it corresponds to the first or second time point. Then create a contrast as [0 0 0 0 ... 0 0 0 1], where there are as many 0s as subjects. The last "1" will correspond to the Time Point variable.

2. Alternatively, if you are using mris_preproc, you can specify

"--paired-diff", then just use a one-sample group mean test. See the --help. This may be easier, and will also allow you to add a between subject factor.

doug

Fornito, Alexander wrote:

...

Hi, I'd like to run a simple repeated-measures analysis tetsing for change in thickness measures between 2 time points. There is no between-subjects factor, only the effect of time. Do I just set up a standard DOSS FSGD file, or do I use a different format? The tutorial only covers between-subjects designs. Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

That did the trick, but I'm having trouble displaying the results. First issue is that I can't seem to load my average surface in tksurfer from the command line.I get the following:

tksurfer average rh inflated surfer: current subjects dir: /data/nelson/work/alex/freesurfer/subjects_twins surfer: not in "scripts" dir ==> using cwd for session root surfer: session root data dir ($session) set to: surfer: /data/nelson/work/alex/freesurfer/subjects_twins ERROR: cound not stat /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1/COR-.info. Does it exist? mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgh, -1): could not open file zcat: /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz.gz: No such file or directory mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz, -1): read error surfer: ### Could not find anatomical header information.

What I do instead is load a subject's surface and then click load > Surface, and select the average one. When I load the sigt file as an overlay, what should I specify as the registration matrix? Clicking "calculate identity matrix" seems to work, so I'm assuming this is ok?

After doing this, I can visualize the SPM ok, but have trouble when I try to obtain scatterplots for a gven vertex. tksurfer closes and I get the error below.(I've tried specifying markers and colors in the FSGD but it doesn't seem to make a difference) I'm using recon-all-nmr,v 1.104 2005/09/07.

gdfReadHeader: reading /data/nelson/work/alex/freesurfer/subjects_twins/stats/twin_qualcheck_fsgd.txt WARNING: Marker for class class_subj1 was invalid. WARNING: Color for class class_subj1 was invalid. WARNING: Marker for class class_subj2 was invalid. WARNING: Color for class class_subj2 was invalid. WARNING: Marker for class class_subj3 was invalid. WARNING: Color for class class_subj3 was invalid. WARNING: Marker for class class_subj4 was invalid. WARNING: Color for class class_subj4 was invalid. WARNING: Marker for class class_subj5 was invalid. WARNING: Color for class class_subj5 was invalid. surfer: dmin=0.8701, vno=132952, x=14.6850, y=-30.1178, z=47.7694 % surfer: curv=0.000000, fs=0.000000 % surfer: val=2.132707, val2=0.000000 % surfer: amp=2.132707, angle=0.000000 deg (0.000000)

Thanks again, A

Oh, I'm sorry. I see the problem now. When I said to create a class for each subject, I meant for each person, not for each FreeSurfer subject. So there should only be 5 classes, not 10.

So do something like this:

Class class_subj1 Class class_subj2 Class class_subj3 Class class_subj4 Class class_subj5

Variables Time Input time1_subj1 class_subj1 -1 Input time2_subj1 class_subj1 +1 ...

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

I still get the same error message. Maybe I've coded the variable wrong?

gdfReadHeader: reading ./test_fsgd.txt INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 11:43 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] Repeated-measures in mris_glm

Yes, mris_preproc is new with stable. In your FSGD file, try specifying the title as a single string without spaces.

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

...

Hi Doug, Is mri_preproc a new addition to the stable version? I'm using recon-all-nmr,v 1.104 2005/09/07 and can't seem to find it. I tried running the analysis through mris_glm as per option 1 as you described, but came across a problem. Command, error, and my FSGD are below. Thanks, Alex

Command:

mris_glm --surfmeas thickness --hemi lh --trgsubj ../average --fsgd ./test_fsgd.txt doss --beta ./beta_doss_thickness-100_lh bfloat --var ./var_doss_thickness-100_lh.w paint --y ./y_doss_thickness-100_lh_000.bfloat --nsmooth 100 gdfReadHeader: reading ./test_fsgd.txt INFO: ignoring tag 1 INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

My FSGD:

GroupDescriptorFile 1 Title Time 1 vs Time 2 Class class_time1_subj1 Class class_time1_subj2 Class class_time1_subj3 Class class_time1_subj4 Class class_time1_subj5 Class class_time2_subj1 Class class_time2_subj2 Class class_time2_subj3 Class class_time2_subj4 Class class_time2_subj5 Variables Time Input time1_subj1 class_time1_subj1 -1 Input time1_subj2 class_time1_subj2 -1 Input time1_subj3 class_time1_subj3 -1 Input time1_subj4 class_time1_subj4 -1 Input time1_subj5 class_time1_subj5 -1 Input time2_subj1 class_time2_subj1 1 Input time2_subj2 class_time2_subj2 1 Input time2_subj3 class_time2_subj3 1 Input time2_subj4 class_time2_subj4 1 Input time2_subj5 class_time2_subj5 1 DefaultVariable Time

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 9:31 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

There are two ways to do it:

1. You can use DOSS, making each subject it's own class. Then create one

"Variable". For a given input, set the variable to either +1 or -1 depending upon whether it corresponds to the first or second time point. Then create a contrast as [0 0 0 0 ... 0 0 0 1], where there are as many 0s as subjects. The last "1" will correspond to the Time Point variable.

2. Alternatively, if you are using mris_preproc, you can specify

"--paired-diff", then just use a one-sample group mean test. See the --help. This may be easier, and will also allow you to add a between subject factor.

doug

Fornito, Alexander wrote:

...

Hi, I'd like to run a simple repeated-measures analysis tetsing for change in thickness measures between 2 time points. There is no between-subjects factor, only the effect of time. Do I just set up a standard DOSS FSGD file, or do I use a different format? The tutorial only covers between-subjects designs. Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

I craeted the average using make_average_surface. An average/mri/T1 directory gets created, but nothing is in there. There is no T1 volume at all. Is this because I'm using an older version of recon-all (v 1.104 2005/09/07)? Could this also explain my trouble creating scatterplots?

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Wed 4/12/2006 2:42 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

How did you create the average subject? Is there a T1.mgz in the mri dir? If so, you might have an old version of tksurfer that is looking for COR files instead of mgz.

Fornito, Alexander wrote:

That did the trick, but I'm having trouble displaying the results. First issue is that I can't seem to load my average surface in tksurfer from the command line.I get the following:

tksurfer average rh inflated surfer: current subjects dir: /data/nelson/work/alex/freesurfer/subjects_twins surfer: not in "scripts" dir ==> using cwd for session root surfer: session root data dir ($session) set to: surfer: /data/nelson/work/alex/freesurfer/subjects_twins ERROR: cound not stat /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1/COR-.info. Does it exist? mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgh, -1): could not open file zcat: /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz.gz: No such file or directory mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz, -1): read error surfer: ### Could not find anatomical header information.

What I do instead is load a subject's surface and then click load > Surface, and select the average one. When I load the sigt file as an overlay, what should I specify as the registration matrix? Clicking "calculate identity matrix" seems to work, so I'm assuming this is ok?

After doing this, I can visualize the SPM ok, but have trouble when I try to obtain scatterplots for a gven vertex. tksurfer closes and I get the error below.(I've tried specifying markers and colors in the FSGD but it doesn't seem to make a difference) I'm using recon-all-nmr,v 1.104 2005/09/07.

gdfReadHeader: reading /data/nelson/work/alex/freesurfer/subjects_twins/stats/twin_qualcheck_fsgd.txt WARNING: Marker for class class_subj1 was invalid. WARNING: Color for class class_subj1 was invalid. WARNING: Marker for class class_subj2 was invalid. WARNING: Color for class class_subj2 was invalid. WARNING: Marker for class class_subj3 was invalid. WARNING: Color for class class_subj3 was invalid. WARNING: Marker for class class_subj4 was invalid. WARNING: Color for class class_subj4 was invalid. WARNING: Marker for class class_subj5 was invalid. WARNING: Color for class class_subj5 was invalid. surfer: dmin=0.8701, vno=132952, x=14.6850, y=-30.1178, z=47.7694 % surfer: curv=0.000000, fs=0.000000 % surfer: val=2.132707, val2=0.000000 % surfer: amp=2.132707, angle=0.000000 deg (0.000000)

Thanks again, A

Oh, I'm sorry. I see the problem now. When I said to create a class for each subject, I meant for each person, not for each FreeSurfer subject. So there should only be 5 classes, not 10.

So do something like this:

Class class_subj1 Class class_subj2 Class class_subj3 Class class_subj4 Class class_subj5

Variables Time Input time1_subj1 class_subj1 -1 Input time2_subj1 class_subj1 +1 ...

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

...

I still get the same error message. Maybe I've coded the variable wrong?

gdfReadHeader: reading ./test_fsgd.txt INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 11:43 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] Repeated-measures in mris_glm

Yes, mris_preproc is new with stable. In your FSGD file, try specifying the title as a single string without spaces.

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

...

Hi Doug, Is mri_preproc a new addition to the stable version? I'm using recon-all-nmr,v 1.104 2005/09/07 and can't seem to find it. I tried running the analysis through mris_glm as per option 1 as you described, but came across a problem. Command, error, and my FSGD are below. Thanks, Alex

Command:

mris_glm --surfmeas thickness --hemi lh --trgsubj ../average --fsgd ./test_fsgd.txt doss --beta ./beta_doss_thickness-100_lh bfloat --var ./var_doss_thickness-100_lh.w paint --y ./y_doss_thickness-100_lh_000.bfloat --nsmooth 100 gdfReadHeader: reading ./test_fsgd.txt INFO: ignoring tag 1 INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

My FSGD:

GroupDescriptorFile 1 Title Time 1 vs Time 2 Class class_time1_subj1 Class class_time1_subj2 Class class_time1_subj3 Class class_time1_subj4 Class class_time1_subj5 Class class_time2_subj1 Class class_time2_subj2 Class class_time2_subj3 Class class_time2_subj4 Class class_time2_subj5 Variables Time Input time1_subj1 class_time1_subj1 -1 Input time1_subj2 class_time1_subj2 -1 Input time1_subj3 class_time1_subj3 -1 Input time1_subj4 class_time1_subj4 -1 Input time1_subj5 class_time1_subj5 -1 Input time2_subj1 class_time2_subj1 1 Input time2_subj2 class_time2_subj2 1 Input time2_subj3 class_time2_subj3 1 Input time2_subj4 class_time2_subj4 1 Input time2_subj5 class_time2_subj5 1 DefaultVariable Time

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 9:31 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

There are two ways to do it:

1. You can use DOSS, making each subject it's own class. Then create one

"Variable". For a given input, set the variable to either +1 or -1 depending upon whether it corresponds to the first or second time point. Then create a contrast as [0 0 0 0 ... 0 0 0 1], where there are as many 0s as subjects. The last "1" will correspond to the Time Point variable.

2. Alternatively, if you are using mris_preproc, you can specify

"--paired-diff", then just use a one-sample group mean test. See the --help. This may be easier, and will also allow you to add a between subject factor.

doug

Fornito, Alexander wrote:

...

Hi, I'd like to run a simple repeated-measures analysis tetsing for change in thickness measures between 2 time points. There is no between-subjects factor, only the effect of time. Do I just set up a standard DOSS FSGD file, or do I use a different format? The tutorial only covers between-subjects designs. Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

make_average_subject did the trick. However, although I can load the lh results as an overlay in tksurfer, if I try to adjust the display settings through View> Configure > Overlay, tksurfer crashes and I get the message below. I don't have this problem when overlaying the rh results.

tksurfer average lh smoothwm surfer: current subjects dir: /data/nelson/work/alex/freesurfer/subjects_twins surfer: not in "scripts" dir ==> using cwd for session root surfer: session root data dir ($session) set to: surfer: /data/nelson/work/alex/freesurfer/subjects_twins ERROR: cound not stat /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1/COR-.info. Does it exist? mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgh, -1): could not open file surfer: Reading header info from /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz surfer: vertices=163842, faces=327680 corRead(): can't open file /data/nelson/work/alex/freesurfer/subjects_twins/average/surf/../mri/orig/COR-.info MRISreadVertexPosition(/data/nelson/work/alex/freesurfer/subjects_twins/average/surf/lh.orig): could not open file /data/nelson/work/alex/freesurfer/subjects_twins/average/surf/lh.orig No such file or directory MRISreadOriginalProperties: could not read surface file /data/nelson/work/alex/freesurfer/subjects_twins/average/surf/lh.orig No such file or directory surfer: single buffered window Xlib: extension "XFree86-DRI" missing on display ":0.0". surfer: using interface /usr/local/freesurfer-dev20050905/lib/tcl/tksurfer.tcl Reading /usr/local/freesurfer-dev20050905/lib/tcl/tkm_common.tcl Reading /usr/local/freesurfer-dev20050905/lib/tcl/tkm_wrappers.tcl Reading /usr/local/freesurfer-dev20050905/lib/tcl/fsgdfPlot.tcl Reading /usr/local/freesurfer-dev20050905/lib/tcl/tkUtils.tcl Read /usr/local/freesurfer-dev20050905/surface_labels.txt, found 95 structures Successfully parsed tksurfer.tcl % offset is 0.25 surfer: reading canonical coordinates from surfer: /data/nelson/work/alex/freesurfer/subjects_twins/average/surf/lh.sphere.reg too big, x major ticks can't be 80000002 Abort

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Thu 4/13/2006 1:45 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

Try using make_average_subject. That will create a T1 volume. Alternatively, you can run make_average_volume, which will just do the volume (make_average_subject will re-do the volume). Or even faster would be to copy the T1 volume from average7 or talairach into your average subject mri dir. You can do this and see if the problem goes away.

doug

Fornito, Alexander wrote:

I craeted the average using make_average_surface. An average/mri/T1 directory gets created, but nothing is in there. There is no T1 volume at all. Is this because I'm using an older version of recon-all (v 1.104 2005/09/07)? Could this also explain my trouble creating scatterplots?

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Wed 4/12/2006 2:42 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

How did you create the average subject? Is there a T1.mgz in the mri dir? If so, you might have an old version of tksurfer that is looking for COR files instead of mgz.

Fornito, Alexander wrote:

...

That did the trick, but I'm having trouble displaying the results. First issue is that I can't seem to load my average surface in tksurfer from the command line.I get the following:

tksurfer average rh inflated surfer: current subjects dir: /data/nelson/work/alex/freesurfer/subjects_twins surfer: not in "scripts" dir ==> using cwd for session root surfer: session root data dir ($session) set to: surfer: /data/nelson/work/alex/freesurfer/subjects_twins ERROR: cound not stat /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1/COR-.info. Does it exist? mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgh, -1): could not open file zcat: /data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz.gz: No such file or directory mghRead(/data/nelson/work/alex/freesurfer/subjects_twins/average/mri/T1.mgz, -1): read error surfer: ### Could not find anatomical header information.

What I do instead is load a subject's surface and then click load > Surface, and select the average one. When I load the sigt file as an overlay, what should I specify as the registration matrix? Clicking "calculate identity matrix" seems to work, so I'm assuming this is ok?

After doing this, I can visualize the SPM ok, but have trouble when I try to obtain scatterplots for a gven vertex. tksurfer closes and I get the error below.(I've tried specifying markers and colors in the FSGD but it doesn't seem to make a difference) I'm using recon-all-nmr,v 1.104 2005/09/07.

gdfReadHeader: reading /data/nelson/work/alex/freesurfer/subjects_twins/stats/twin_qualcheck_fsgd.txt WARNING: Marker for class class_subj1 was invalid. WARNING: Color for class class_subj1 was invalid. WARNING: Marker for class class_subj2 was invalid. WARNING: Color for class class_subj2 was invalid. WARNING: Marker for class class_subj3 was invalid. WARNING: Color for class class_subj3 was invalid. WARNING: Marker for class class_subj4 was invalid. WARNING: Color for class class_subj4 was invalid. WARNING: Marker for class class_subj5 was invalid. WARNING: Color for class class_subj5 was invalid. surfer: dmin=0.8701, vno=132952, x=14.6850, y=-30.1178, z=47.7694 % surfer: curv=0.000000, fs=0.000000 % surfer: val=2.132707, val2=0.000000 % surfer: amp=2.132707, angle=0.000000 deg (0.000000)

Thanks again, A

Oh, I'm sorry. I see the problem now. When I said to create a class for each subject, I meant for each person, not for each FreeSurfer subject. So there should only be 5 classes, not 10.

So do something like this:

Class class_subj1 Class class_subj2 Class class_subj3 Class class_subj4 Class class_subj5

Variables Time Input time1_subj1 class_subj1 -1 Input time2_subj1 class_subj1 +1 ...

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

...

I still get the same error message. Maybe I've coded the variable wrong?

gdfReadHeader: reading ./test_fsgd.txt INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 11:43 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: RE: [Freesurfer] Repeated-measures in mris_glm

Yes, mris_preproc is new with stable. In your FSGD file, try specifying the title as a single string without spaces.

doug

On Tue, 11 Apr 2006, Fornito, Alexander wrote:

...

Hi Doug, Is mri_preproc a new addition to the stable version? I'm using recon-all-nmr,v 1.104 2005/09/07 and can't seem to find it. I tried running the analysis through mris_glm as per option 1 as you described, but came across a problem. Command, error, and my FSGD are below. Thanks, Alex

Command:

mris_glm --surfmeas thickness --hemi lh --trgsubj ../average --fsgd ./test_fsgd.txt doss --beta ./beta_doss_thickness-100_lh bfloat --var ./var_doss_thickness-100_lh.w paint --y ./y_doss_thickness-100_lh_000.bfloat --nsmooth 100 gdfReadHeader: reading ./test_fsgd.txt INFO: ignoring tag 1 INFO: gd2mtx_method is doss ERROR: Design Matrix: nrows (10) <= ncols (11)

My FSGD:

GroupDescriptorFile 1 Title Time 1 vs Time 2 Class class_time1_subj1 Class class_time1_subj2 Class class_time1_subj3 Class class_time1_subj4 Class class_time1_subj5 Class class_time2_subj1 Class class_time2_subj2 Class class_time2_subj3 Class class_time2_subj4 Class class_time2_subj5 Variables Time Input time1_subj1 class_time1_subj1 -1 Input time1_subj2 class_time1_subj2 -1 Input time1_subj3 class_time1_subj3 -1 Input time1_subj4 class_time1_subj4 -1 Input time1_subj5 class_time1_subj5 -1 Input time2_subj1 class_time2_subj1 1 Input time2_subj2 class_time2_subj2 1 Input time2_subj3 class_time2_subj3 1 Input time2_subj4 class_time2_subj4 1 Input time2_subj5 class_time2_subj5 1 DefaultVariable Time

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

-----Original Message----- From: Doug Greve [mailto:greve@nmr.mgh.harvard.edu] Sent: Tue 4/11/2006 9:31 AM To: Fornito, Alexander Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Repeated-measures in mris_glm

There are two ways to do it:

1. You can use DOSS, making each subject it's own class. Then create one

"Variable". For a given input, set the variable to either +1 or -1 depending upon whether it corresponds to the first or second time point. Then create a contrast as [0 0 0 0 ... 0 0 0 1], where there are as many 0s as subjects. The last "1" will correspond to the Time Point variable.

2. Alternatively, if you are using mris_preproc, you can specify

"--paired-diff", then just use a one-sample group mean test. See the --help. This may be easier, and will also allow you to add a between subject factor.

doug

Fornito, Alexander wrote:

...

Hi, I'd like to run a simple repeated-measures analysis tetsing for change in thickness measures between 2 time points. There is no between-subjects factor, only the effect of time. Do I just set up a standard DOSS FSGD file, or do I use a different format? The tutorial only covers between-subjects designs. Thanks, Alex

Alex Fornito M.Psych/PhD (clin. neuro.) candidate Melbourne Neuropsychiatry Centre and Department of Psychology National Neuroscience Facility The University of Melbourne Levels 2 & 3, Alan Gilbert Building 161 Barry St Carlton South Vic 3053 Australia Ph: +61 3 8344 1624 Fax: +61 3 9348 0469 email: alexander.fornito@wh.org.au

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer