Hi,
I have to bring up an issue again which we had some discussion about before, but which we haven't yet found a satisfying solution. I would like pose my question a little more precisely again: 1) I have a dataset that is in 181x217x181 coordinates. 2) I am processing it as follows:
mri_convert name.img name/mri/orig/001.mgz recon-all -motioncor -subjid name
3) after step 2, the resulting dataset is somehow transformed to a 256x256x256 dataset, i. e. the 181x217x181 is embedded into the 256^3 array: it is rotated and the origin in 181x217x181 is not the origin in the 256x256x256 array. 4) Since I would like to do a voxelwise comparison between the recon-all result and my initial dataset 1), I need to backtransform the result into the 181x217x181 coordinates (and of course, the backtransformation should be the exact inverse of the transformation in 2). I haven't found any info about how this transformation is done by FS, mri_info does not provide it.
Thanks, Frederick
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Hi Frederick, freesurfer always transforms your input data into 256 cubed. It does, however, store information on this transform, so you can get your data back into the original space if you want to. Let's say you wanted to get Freesurfer's segmentation back into original space. You could cd into your subject's mri subdirectory and do: mri_convert -rl rawavg.mgz -rt nearest aseg.mgz aseg_in_raw_space.mgz Cheers, Johannes
This should do it:
tkregister2 --targ rawavg.mgz --mov orig.mgz --reg reg.dat --noedit mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
If this does not work with your version of vol2vol, then use the one here:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_vol2vol
Frederick Klauschen wrote:
Hi,
I have to bring up an issue again which we had some discussion about before, but which we haven't yet found a satisfying solution. I would like pose my question a little more precisely again:
- I have a dataset that is in 181x217x181
coordinates. 2) I am processing it as follows:
mri_convert name.img name/mri/orig/001.mgz recon-all -motioncor -subjid name
- after step 2, the resulting dataset is somehow
transformed to a 256x256x256 dataset, i. e. the 181x217x181 is embedded into the 256^3 array: it is rotated and the origin in 181x217x181 is not the origin in the 256x256x256 array. 4) Since I would like to do a voxelwise comparison between the recon-all result and my initial dataset 1), I need to backtransform the result into the 181x217x181 coordinates (and of course, the backtransformation should be the exact inverse of the transformation in 2). I haven't found any info about how this transformation is done by FS, mri_info does not provide it.
Thanks, Frederick
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Doug,
I am running Freesurfer on MacOsX Darwin. The command as you suggested: mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
using the mri_vol2vol version that came with FS I get the message: ERROR: Option --targ unknown
using the mri_vol2vol from the link you provided results in: Exec format error. Binary file not executable. Probably, that is a version compiled for x86 and not Mac?
Thanks for your help, Frederick
--- Doug Greve greve@nmr.mgh.harvard.edu wrote:
This should do it:
tkregister2 --targ rawavg.mgz --mov orig.mgz --reg reg.dat --noedit mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
If this does not work with your version of vol2vol, then use the one here:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_vol2vol
Frederick Klauschen wrote:
Hi,
I have to bring up an issue again which we had some discussion about before, but which we haven't yet found a satisfying solution. I would like pose my
question a
little more precisely again:
- I have a dataset that is in 181x217x181
coordinates. 2) I am processing it as follows:
mri_convert name.img name/mri/orig/001.mgz recon-all -motioncor -subjid name
- after step 2, the resulting dataset is somehow
transformed to a 256x256x256 dataset, i. e. the 181x217x181 is embedded into the 256^3 array: it is rotated and the origin in 181x217x181 is not
the
origin in the 256x256x256 array. 4) Since I would like to do a voxelwise comparison between the recon-all result and my initial dataset 1), I need to backtransform the result into the 181x217x181 coordinates (and of course, the backtransformation should be the exact inverse of
the
transformation in 2). I haven't found any info
about
how this transformation is done by FS, mri_info
does
not provide it.
Thanks, Frederick
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam
protection around
http://mail.yahoo.com _______________________________________________ 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 greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Ah, yes, that one was for linux. I've copied a mac version there now (has .mac extension).
doug
Frederick Klauschen wrote:
Hi Doug,
I am running Freesurfer on MacOsX Darwin. The command as you suggested: mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
using the mri_vol2vol version that came with FS I get the message: ERROR: Option --targ unknown
using the mri_vol2vol from the link you provided results in: Exec format error. Binary file not executable. Probably, that is a version compiled for x86 and not Mac?
Thanks for your help, Frederick
--- Doug Greve greve@nmr.mgh.harvard.edu wrote:
This should do it:
tkregister2 --targ rawavg.mgz --mov orig.mgz --reg reg.dat --noedit mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
If this does not work with your version of vol2vol, then use the one here:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_vol2vol
Frederick Klauschen wrote:
Hi,
I have to bring up an issue again which we had some discussion about before, but which we haven't yet found a satisfying solution. I would like pose my
question a
little more precisely again:
- I have a dataset that is in 181x217x181
coordinates. 2) I am processing it as follows:
mri_convert name.img name/mri/orig/001.mgz recon-all -motioncor -subjid name
- after step 2, the resulting dataset is somehow
transformed to a 256x256x256 dataset, i. e. the 181x217x181 is embedded into the 256^3 array: it is rotated and the origin in 181x217x181 is not
the
origin in the 256x256x256 array. 4) Since I would like to do a voxelwise comparison between the recon-all result and my initial dataset 1), I need to backtransform the result into the 181x217x181 coordinates (and of course, the backtransformation should be the exact inverse of
the
transformation in 2). I haven't found any info
about
how this transformation is done by FS, mri_info
does
not provide it.
Thanks, Frederick
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam
protection around
http://mail.yahoo.com _______________________________________________ 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 greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Hi Doug,
when I run the following: mri_vol2vol.mac --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o originraw.mgz
I get this error message:
dyld: Library not loaded: libvnl_algo.dylib Referenced from: /Applications/freesurfer/bin//mri_vol2vol.mac Reason: image not found Trace/BPT trap
I don't know if the reason is that the library is not found or that it doesn't find the image file, which would be strange because all files in in the correct path. I couldn't find libvnl_algo.dylib. (I use the latest FS version)
Frederick
--- Doug Greve greve@nmr.mgh.harvard.edu wrote:
Ah, yes, that one was for linux. I've copied a mac version there now (has .mac extension).
doug
Frederick Klauschen wrote:
Hi Doug,
I am running Freesurfer on MacOsX Darwin. The
command
as you suggested: mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
using the mri_vol2vol version that came with FS I
get
the message: ERROR: Option --targ unknown
using the mri_vol2vol from the link you provided results in: Exec format error. Binary file not executable. Probably, that is a version compiled for x86 and
not
Mac?
Thanks for your help, Frederick
--- Doug Greve greve@nmr.mgh.harvard.edu wrote:
This should do it:
tkregister2 --targ rawavg.mgz --mov orig.mgz
--reg
reg.dat --noedit mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
If this does not work with your version of
vol2vol,
then use the one here:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_vol2vol
Frederick Klauschen wrote:
Hi,
I have to bring up an issue again which we had
some
discussion about before, but which we haven't yet found a satisfying solution. I would like pose my
question a
little more precisely again:
- I have a dataset that is in 181x217x181
coordinates. 2) I am processing it as follows:
mri_convert name.img name/mri/orig/001.mgz recon-all -motioncor -subjid name
- after step 2, the resulting dataset is somehow
transformed to a 256x256x256 dataset, i. e. the 181x217x181 is embedded into the 256^3 array: it is rotated and the origin in 181x217x181 is
not
the
origin in the 256x256x256 array. 4) Since I would like to do a voxelwise
comparison
between the recon-all result and my initial
dataset
1), I need to backtransform the result into the 181x217x181 coordinates (and of course, the backtransformation should be the exact inverse of
the
transformation in 2). I haven't found any info
about
how this transformation is done by FS, mri_info
does
not provide it.
Thanks, Frederick
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam
protection around
http://mail.yahoo.com _______________________________________________ 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 greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam
protection around
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Frederick,
I've put the file:
vnl-dylib-mac.tar.gz
here:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/fsdev/nicks
Copy it to the same directory as your mri_vol2vol, then type:
tar zxvf vnl-dylib-mac.tar.gz
and the necessary VXL libs will exist.
Nick
On Wed, 2006-09-13 at 11:29 -0700, Frederick Klauschen wrote:
Hi Doug,
when I run the following: mri_vol2vol.mac --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o originraw.mgz
I get this error message:
dyld: Library not loaded: libvnl_algo.dylib Referenced from: /Applications/freesurfer/bin//mri_vol2vol.mac Reason: image not found Trace/BPT trap
I don't know if the reason is that the library is not found or that it doesn't find the image file, which would be strange because all files in in the correct path. I couldn't find libvnl_algo.dylib. (I use the latest FS version)
Frederick
--- Doug Greve greve@nmr.mgh.harvard.edu wrote:
Ah, yes, that one was for linux. I've copied a mac version there now (has .mac extension).
doug
Frederick Klauschen wrote:
Hi Doug,
I am running Freesurfer on MacOsX Darwin. The
command
as you suggested: mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
using the mri_vol2vol version that came with FS I
get
the message: ERROR: Option --targ unknown
using the mri_vol2vol from the link you provided results in: Exec format error. Binary file not executable. Probably, that is a version compiled for x86 and
not
Mac?
Thanks for your help, Frederick
--- Doug Greve greve@nmr.mgh.harvard.edu wrote:
This should do it:
tkregister2 --targ rawavg.mgz --mov orig.mgz
--reg
reg.dat --noedit mri_vol2vol --targ rawavg.mgz --mov orig.mgz --reg reg.dat --o orig-in-rawavg.mgz
If this does not work with your version of
vol2vol,
then use the one here:
ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/mri_vol2vol
Frederick Klauschen wrote:
Hi,
I have to bring up an issue again which we had
some
discussion about before, but which we haven't yet found a satisfying solution. I would like pose my
question a
little more precisely again:
- I have a dataset that is in 181x217x181
coordinates. 2) I am processing it as follows:
mri_convert name.img name/mri/orig/001.mgz recon-all -motioncor -subjid name
- after step 2, the resulting dataset is somehow
transformed to a 256x256x256 dataset, i. e. the 181x217x181 is embedded into the 256^3 array: it is rotated and the origin in 181x217x181 is
not
the
origin in the 256x256x256 array. 4) Since I would like to do a voxelwise
comparison
between the recon-all result and my initial
dataset
1), I need to backtransform the result into the 181x217x181 coordinates (and of course, the backtransformation should be the exact inverse of
the
transformation in 2). I haven't found any info
about
how this transformation is done by FS, mri_info
does
not provide it.
Thanks, Frederick
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam
protection around
http://mail.yahoo.com _______________________________________________ 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 greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam
protection around
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
In order to help us help you, please follow the steps in: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting
Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Doug Greve -
Frederick Klauschen -
Johannes Klein -
Nick Schmansky