Hi,
We are trying to extract certain regions from FreeSurfer segmentation. There are couple of regions which are not found in the output files. We are currently searching for the regions in 13 output files. The regions are the following: right fornix (508), left fornix (558), right PLIC (158), left PLIC (157) and the starting, way and end points for cingulum and CST.
Could you let us know in which file would we be able to find these regions? Is it possible that some regions are empty?
Thank you for your help
Judit Haasz
Hi Judit,
the fornix is labeled in the aseg, not the parcellation, although this is not the default). If you rerun mri_cc with the -fornix option it will write a volume named mri/fornix.mgz with the fornix in it.
cheers, Bruce
On Wed, 1 Jul 2009, Judit Haasz wrote:
Hi,
We are trying to extract certain regions from FreeSurfer segmentation. There are couple of regions which are not found in the output files. We are currently searching for the regions in 13 output files. The regions are the following: right fornix (508), left fornix (558), right PLIC (158), left PLIC (157) and the starting, way and end points for cingulum and CST.
Could you let us know in which file would we be able to find these regions? Is it possible that some regions are empty?
Thank you for your help
Judit Haasz _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Bruce Fischl wrote:
Hi Judit,
the fornix is labeled in the aseg, not the parcellation, although this is not the default). If you rerun mri_cc with the -fornix option it will write a volume named mri/fornix.mgz with the fornix in it.
cheers, Bruce
On Wed, 1 Jul 2009, Judit Haasz wrote:
Hi,
We are trying to extract certain regions from FreeSurfer segmentation. There are couple of regions which are not found in the output files. We are currently searching for the regions in 13 output files. The regions are the following: right fornix (508), left fornix (558), right PLIC (158), left PLIC (157) and the starting, way and end points for cingulum and CST.
Could you let us know in which file would we be able to find these regions? Is it possible that some regions are empty?
Thank you for your help
Judit Haasz _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Bruce and all,
I tried using the command as you suggested, but i get the following message:
mri_cc -fornix STROKE_3T including fornix in segmentation reading aseg from /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz mri_cc: volume /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz already has CC in it
Apart from that I encountered two more problems. Our subject group consists of 115 healthy elderly people. Some of the subjects` scan exhibit intensified white matter lesion which results in inaccurate pial surface reconstruction, that is pial surface penentrates into the white matter. Is there any way we could overcome this problem and obtain a more reliable segmentation? My next question refers to the problem of different FreeSurfer versions. We have data from 2005 and 2009. The ones from 2005 were processed by FreeSurfer ver 3.01. Now we are using FreeSurfer 4.3.0 with a separate mri_normalize program, that we received from you since a systematic error in the segmentation process was present. We are planning to compare the two datasets. We are aware of the fact that we have to rerun all subjects from 2005 with the version we are currently using. I am just wondering if there is a way to keep our previous wm and pial surface edits. As far as i know, edits can be kept if FreeSurfer is run from autorecon2. Would it be an option for us even though, mri_normalize is involved in autorecon1.
I really appreciate your help
Judit Haasz
Hi Judit,
all the edits will be kept, so that shouldn't be a problem. It's hard to diagnose the wm segmentation without looking at a volume, but if they are really hyperintense you may need to manually fill them in in the wm.mgz. As for mri_cc, you should use aseg.auto_noCCseg.mgz instead of aseg.mgz via:
mri_cc -aseg auto_noCCseg ...
what is PLIC? The fornix should be non-zero after running this command.
cheers, Bruce
On Mon, 20 Jul 2009, Judit Haasz wrote:
Bruce Fischl wrote:
Hi Judit,
the fornix is labeled in the aseg, not the parcellation, although this is not the default). If you rerun mri_cc with the -fornix option it will write a volume named mri/fornix.mgz with the fornix in it.
cheers, Bruce
On Wed, 1 Jul 2009, Judit Haasz wrote:
Hi,
We are trying to extract certain regions from FreeSurfer segmentation. There are couple of regions which are not found in the output files. We are currently searching for the regions in 13 output files. The regions are the following: right fornix (508), left fornix (558), right PLIC (158), left PLIC (157) and the starting, way and end points for cingulum and CST.
Could you let us know in which file would we be able to find these regions? Is it possible that some regions are empty?
Thank you for your help
Judit Haasz _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Bruce and all,
I tried using the command as you suggested, but i get the following message:
mri_cc -fornix STROKE_3T including fornix in segmentation reading aseg from /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz mri_cc: volume /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz already has CC in it
Apart from that I encountered two more problems. Our subject group consists of 115 healthy elderly people. Some of the subjects` scan exhibit intensified white matter lesion which results in inaccurate pial surface reconstruction, that is pial surface penentrates into the white matter. Is there any way we could overcome this problem and obtain a more reliable segmentation? My next question refers to the problem of different FreeSurfer versions. We have data from 2005 and 2009. The ones from 2005 were processed by FreeSurfer ver 3.01. Now we are using FreeSurfer 4.3.0 with a separate mri_normalize program, that we received from you since a systematic error in the segmentation process was present. We are planning to compare the two datasets. We are aware of the fact that we have to rerun all subjects from 2005 with the version we are currently using. I am just wondering if there is a way to keep our previous wm and pial surface edits. As far as i know, edits can be kept if FreeSurfer is run from autorecon2. Would it be an option for us even though, mri_normalize is involved in autorecon1.
I really appreciate your help
Judit Haasz
Bruce Fischl wrote:
Hi Judit,
all the edits will be kept, so that shouldn't be a problem. It's hard to diagnose the wm segmentation without looking at a volume, but if they are really hyperintense you may need to manually fill them in in the wm.mgz. As for mri_cc, you should use aseg.auto_noCCseg.mgz instead of aseg.mgz via:
mri_cc -aseg auto_noCCseg ...
what is PLIC? The fornix should be non-zero after running this command.
cheers, Bruce
On Mon, 20 Jul 2009, Judit Haasz wrote:
Bruce Fischl wrote:
Hi Judit,
the fornix is labeled in the aseg, not the parcellation, although this is not the default). If you rerun mri_cc with the -fornix option it will write a volume named mri/fornix.mgz with the fornix in it.
cheers, Bruce
On Wed, 1 Jul 2009, Judit Haasz wrote:
Hi,
We are trying to extract certain regions from FreeSurfer segmentation. There are couple of regions which are not found in the output files. We are currently searching for the regions in 13 output files. The regions are the following: right fornix (508), left fornix (558), right PLIC (158), left PLIC (157) and the starting, way and end points for cingulum and CST.
Could you let us know in which file would we be able to find these regions? Is it possible that some regions are empty?
Thank you for your help
Judit Haasz _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Bruce and all,
I tried using the command as you suggested, but i get the following message:
mri_cc -fornix STROKE_3T including fornix in segmentation reading aseg from /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz mri_cc: volume /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz already has CC in it
Apart from that I encountered two more problems. Our subject group consists of 115 healthy elderly people. Some of the subjects` scan exhibit intensified white matter lesion which results in inaccurate pial surface reconstruction, that is pial surface penentrates into the white matter. Is there any way we could overcome this problem and obtain a more reliable segmentation? My next question refers to the problem of different FreeSurfer versions. We have data from 2005 and 2009. The ones from 2005 were processed by FreeSurfer ver 3.01. Now we are using FreeSurfer 4.3.0 with a separate mri_normalize program, that we received from you since a systematic error in the segmentation process was present. We are planning to compare the two datasets. We are aware of the fact that we have to rerun all subjects from 2005 with the version we are currently using. I am just wondering if there is a way to keep our previous wm and pial surface edits. As far as i know, edits can be kept if FreeSurfer is run from autorecon2. Would it be an option for us even though, mri_normalize is involved in autorecon1.
I really appreciate your help
Judit Haasz
Hi,
mri_cc -fornix works now. Is it possible to segment any other white matter pathway or tracking point that are not in the default, but appear in FreeSurferColorLut? Considering pial surface reconstruction, what can i do if the pial surface penentrates into the cortical GM, not following the brain surface? How shall I proceed with re-running subjects from 2005? How would all of our edits be kept?
Thanks, Judit
Hi Judit, there is a white matter parcellation that is run by default called wmparc.mgz. Anything else is future work. The pial surface misplacement is usually a result of the white matter being wrong, but the intervention depends on the problem, so we'd need to see the example. As for the edits, they should be kept when you rerun -all.
cheers, Bruce
On Tue, 21 Jul 2009, Judit Haasz wrote:
Bruce Fischl wrote:
Hi Judit,
all the edits will be kept, so that shouldn't be a problem. It's hard to diagnose the wm segmentation without looking at a volume, but if they are really hyperintense you may need to manually fill them in in the wm.mgz. As for mri_cc, you should use aseg.auto_noCCseg.mgz instead of aseg.mgz via:
mri_cc -aseg auto_noCCseg ...
what is PLIC? The fornix should be non-zero after running this command.
cheers, Bruce
On Mon, 20 Jul 2009, Judit Haasz wrote:
Bruce Fischl wrote:
Hi Judit,
the fornix is labeled in the aseg, not the parcellation, although this is not the default). If you rerun mri_cc with the -fornix option it will write a volume named mri/fornix.mgz with the fornix in it.
cheers, Bruce
On Wed, 1 Jul 2009, Judit Haasz wrote:
Hi,
We are trying to extract certain regions from FreeSurfer segmentation. There are couple of regions which are not found in the output files. We are currently searching for the regions in 13 output files. The regions are the following: right fornix (508), left fornix (558), right PLIC (158), left PLIC (157) and the starting, way and end points for cingulum and CST.
Could you let us know in which file would we be able to find these regions? Is it possible that some regions are empty?
Thank you for your help
Judit Haasz _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Bruce and all,
I tried using the command as you suggested, but i get the following message:
mri_cc -fornix STROKE_3T including fornix in segmentation reading aseg from /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz mri_cc: volume /Applications/freesurfer//subjects/STROKE_3T/mri/aseg.mgz already has CC in it
Apart from that I encountered two more problems. Our subject group consists of 115 healthy elderly people. Some of the subjects` scan exhibit intensified white matter lesion which results in inaccurate pial surface reconstruction, that is pial surface penentrates into the white matter. Is there any way we could overcome this problem and obtain a more reliable segmentation? My next question refers to the problem of different FreeSurfer versions. We have data from 2005 and 2009. The ones from 2005 were processed by FreeSurfer ver 3.01. Now we are using FreeSurfer 4.3.0 with a separate mri_normalize program, that we received from you since a systematic error in the segmentation process was present. We are planning to compare the two datasets. We are aware of the fact that we have to rerun all subjects from 2005 with the version we are currently using. I am just wondering if there is a way to keep our previous wm and pial surface edits. As far as i know, edits can be kept if FreeSurfer is run from autorecon2. Would it be an option for us even though, mri_normalize is involved in autorecon1.
I really appreciate your help
Judit Haasz
Hi,
mri_cc -fornix works now. Is it possible to segment any other white matter pathway or tracking point that are not in the default, but appear in FreeSurferColorLut? Considering pial surface reconstruction, what can i do if the pial surface penentrates into the cortical GM, not following the brain surface? How shall I proceed with re-running subjects from 2005? How would all of our edits be kept?
Thanks, Judit
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Judit Haasz