Re: [Freesurfer] Improving the accuracy of average surfaces made from make_average_subject
Hi Miggy
can you cc the list so that others can answer? The two atlases we provide are described in:
https://surfer.nmr.mgh.harvard.edu/ftp/articles/fischl04-parcellation.pdf https://surfer.nmr.mgh.harvard.edu/ftp/articles/desikan06-parcellation.pdf http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2937159/pdf/nihms213933.pdf
and yes, I guess you could edit the ribbon.mgz and then convert it back to wm.mgz. You'll need to make sure that the "on" edits are turned to 255 and the "off" ones to 1 (not 0) for things to work properly
cheers Bruce On Wed, 4 Sep 2013, Miggy Chuapoco wrote:
Hi Bruce, Thanks for the help. To explain a bit more about the first issue, we convert ribbon.mgz into a a nifti file using a script in matlab, which we then edit in ITK-SNAP to fix the segmentation. Typically from there we use a program called mrVista to create meshes and do other data analysis depending on what we need to do. However, we are doing group averaging for the first time and we would like to use these fixed segmentations that we have in the group averaging, and not the raw segmentations from recon-all. I was curious to know that if I were able to convert the nifti file that we have edited back into an .mgz file, would we be able to theoretically obtain a new wm.mgz file and run autorecon2-wm. Let me know if that makes sense.
With regards to the second issue, which atlas is used in the make_average_subject process and where would I be able to find a list of the annotation files within each atlas?
Thanks for your help! Miggy
On Wed, Sep 4, 2013 at 1:45 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote: Hi Miggy,
not sure about your first issue as the ribbon is the gray matter and the wm.mgz is white matter, so how would edits in one change the other? For the second one, there is a flag in mris_register that allows you to draw a label on an individual subject and indicate what parcellation unit is should be in the average. The syntax is: mris_register -l <label file> <gcsa file> <annotation name> ... where <label_file> is the label you draw on the individual (e.g. in tksurfer) <gcas_file> is one of our standard atlases, listed in $FREESURFER_HOME/average/*.gcs <annotation name> is the name of the annotation in the atlas that the label should map to (e.g. S_central in the Destrieux atlas) cheers Bruce On Wed, 4 Sep 2013, Miggy Chuapoco wrote: Hi all, I'm trying to make a group average of a set of subjects that we've collected various functional data on and use anatomical landmarks of the group average surface to predict the locations of functional areas that we've found to be correlated to the anatomy of the ventral cortex. As such, we'd like to get as accurate of an average surface as possible. I've come up on two problems that I'd like to address and was wondering if anyone could help come up with a solution. I ran the recon-all command on all the subjects and created a segmentation, however we made edits to the segmentation using ITK-SNAP, and not tkmedit. Some of the changes were fairly drastic and we'd like to use the fixed segmentations in the averaging process. I've considered perhaps running the -autorecon2-wm flag, but unfortunately using ITK-SNAP essentially makes edits to ribbon.mgz, and not wm.mgz. I was wondering if there was a way to "extract" a new wm.mgz from an edited ribbon.mgz. My second problem involves trying to anchor a sulcus on the ventral surface during the averaging process (the mid-fusiform sulcus) in hopes of getting a more accurate representation of the sulcus on the average surface. As it stands there is a bit too much variability of the average MFS location when compared to the individual subjects' MFS location, and we'd like to reduce that if possible. Is there any way of using a label, or something similar, that can be considered as an "anchor" during the averaging process? Thanks and regards, Miggy ChuapocoThe 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.
Hi Bruce,
I was able to change the white matter label values to 255 and the rest to 1 however when I ran -autorecon2-wm the new surface looks the same as the old when I view them both in tksurfer. Should I be running something else with recon-all to get a fixed segmentation?
Miggy
On Wed, Sep 4, 2013 at 2:32 PM, Bruce Fischl fischl@nmr.mgh.harvard.eduwrote:
Hi Miggy
can you cc the list so that others can answer? The two atlases we provide are described in:
https://surfer.nmr.mgh.**harvard.edu/ftp/articles/** fischl04-parcellation.pdfhttps://surfer.nmr.mgh.harvard.edu/ftp/articles/fischl04-parcellation.pdf https://surfer.nmr.mgh.**harvard.edu/ftp/articles/** desikan06-parcellation.pdfhttps://surfer.nmr.mgh.harvard.edu/ftp/articles/desikan06-parcellation.pdf http://www.ncbi.nlm.nih.gov/**pmc/articles/PMC2937159/pdf/** nihms213933.pdfhttp://www.ncbi.nlm.nih.gov/pmc/articles/PMC2937159/pdf/nihms213933.pdf
and yes, I guess you could edit the ribbon.mgz and then convert it back to wm.mgz. You'll need to make sure that the "on" edits are turned to 255 and the "off" ones to 1 (not 0) for things to work properly
cheers Bruce On Wed, 4 Sep 2013, Miggy Chuapoco wrote:
Hi Bruce,
Thanks for the help. To explain a bit more about the first issue, we convert ribbon.mgz into a a nifti file using a script in matlab, which we then edit in ITK-SNAP to fix the segmentation. Typically from there we use a program called mrVista to create meshes and do other data analysis depending on what we need to do. However, we are doing group averaging for the first time and we would like to use these fixed segmentations that we have in the group averaging, and not the raw segmentations from recon-all. I was curious to know that if I were able to convert the nifti file that we have edited back into an .mgz file, would we be able to theoretically obtain a new wm.mgz file and run autorecon2-wm. Let me know if that makes sense.
With regards to the second issue, which atlas is used in the make_average_subject process and where would I be able to find a list of the annotation files within each atlas?
Thanks for your help! Miggy
On Wed, Sep 4, 2013 at 1:45 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote: Hi Miggy,
not sure about your first issue as the ribbon is the gray matter and the wm.mgz is white matter, so how would edits in one change the other? For the second one, there is a flag in mris_register that allows you to draw a label on an individual subject and indicate what parcellation unit is should be in the average. The syntax is: mris_register -l <label file> <gcsa file> <annotation name> ... where <label_file> is the label you draw on the individual (e.g. in tksurfer) <gcas_file> is one of our standard atlases, listed in $FREESURFER_HOME/average/*.gcs <annotation name> is the name of the annotation in the atlas that the label should map to (e.g. S_central in the Destrieux atlas) cheers Bruce On Wed, 4 Sep 2013, Miggy Chuapoco wrote: Hi all, I'm trying to make a group average of a set of subjects that we've collected various functional data on and use anatomical landmarks of the group average surface to predict the locations of functional areas that we've found to be correlated to the anatomy of the ventral cortex. As such, we'd like to get as accurate of an average surface as possible. I've come up on two problems that I'd like to address and was wondering if anyone could help come up with a solution. I ran the recon-all command on all the subjects and created a segmentation, however we made edits to the segmentation using ITK-SNAP, and not tkmedit. Some of the changes were fairly drastic and we'd like to use the fixed segmentations in the averaging process. I've considered perhaps running the -autorecon2-wm flag, but unfortunately using ITK-SNAP essentially makes edits to ribbon.mgz, and not wm.mgz. I was wondering if there was a way to "extract" a new wm.mgz from an edited ribbon.mgz. My second problem involves trying to anchor a sulcus on the ventral surface during the averaging process (the mid-fusiform sulcus) in hopes of getting a more accurate representation of the sulcus on the average surface. As it stands there is a bit too much variability of the average MFS location when compared to the individual subjects' MFS location, and we'd like to reduce that if possible. Is there any way of using a label, or something similar, that can be considered as an "anchor" during the averaging process? Thanks and regards, Miggy ChuapocoThe 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/**compliancelinehttp://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.
you need to run autorecon3 also On Wed, 4 Sep 2013, Miggy Chuapoco wrote:
Hi Bruce, I was able to change the white matter label values to 255 and the rest to 1 however when I ran -autorecon2-wm the new surface looks the same as the old when I view them both in tksurfer. Should I be running something else with recon-all to get a fixed segmentation?
Miggy
On Wed, Sep 4, 2013 at 2:32 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote: Hi Miggy
can you cc the list so that others can answer? The two atlases we provide are described in: https://surfer.nmr.mgh.harvard.edu/ftp/articles/fischl04-parcellation.pdf https://surfer.nmr.mgh.harvard.edu/ftp/articles/desikan06-parcellation.pdf http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2937159/pdf/nihms213933.pdf and yes, I guess you could edit the ribbon.mgz and then convert it back to wm.mgz. You'll need to make sure that the "on" edits are turned to 255 and the "off" ones to 1 (not 0) for things to work properly cheers Bruce On Wed, 4 Sep 2013, Miggy Chuapoco wrote: Hi Bruce, Thanks for the help. To explain a bit more about the first issue, we convert ribbon.mgz into a a nifti file using a script in matlab, which we then edit in ITK-SNAP to fix the segmentation. Typically from there we use a program called mrVista to create meshes and do other data analysis depending on what we need to do. However, we are doing group averaging for the first time and we would like to use these fixed segmentations that we have in the group averaging, and not the raw segmentations from recon-all. I was curious to know that if I were able to convert the nifti file that we have edited back into an .mgz file, would we be able to theoretically obtain a new wm.mgz file and run autorecon2-wm. Let me know if that makes sense. With regards to the second issue, which atlas is used in the make_average_subject process and where would I be able to find a list of the annotation files within each atlas? Thanks for your help! Miggy On Wed, Sep 4, 2013 at 1:45 PM, Bruce Fischl <fischl@nmr.mgh.harvard.edu> wrote: Hi Miggy, not sure about your first issue as the ribbon is the gray matter and the wm.mgz is white matter, so how would edits in one change the other? For the second one, there is a flag in mris_register that allows you to draw a label on an individual subject and indicate what parcellation unit is should be in the average. The syntax is: mris_register -l <label file> <gcsa file> <annotation name> ... where <label_file> is the label you draw on the individual (e.g. in tksurfer) <gcas_file> is one of our standard atlases, listed in $FREESURFER_HOME/average/*.gcs <annotation name> is the name of the annotation in the atlas that the label should map to (e.g. S_central in the Destrieux atlas) cheers Bruce On Wed, 4 Sep 2013, Miggy Chuapoco wrote: Hi all, I'm trying to make a group average of a set of subjects that we've collected various functional data on and use anatomical landmarks of the group average surface to predict the locations of functional areas that we've found to be correlated to the anatomy of the ventral cortex. As such, we'd like to get as accurate of an average surface as possible. I've come up on two problems that I'd like to address and was wondering if anyone could help come up with a solution. I ran the recon-all command on all the subjects and created a segmentation, however we made edits to the segmentation using ITK-SNAP, and not tkmedit. Some of the changes were fairly drastic and we'd like to use the fixed segmentations in the averaging process. I've considered perhaps running the -autorecon2-wm flag, but unfortunately using ITK-SNAP essentially makes edits to ribbon.mgz, and not wm.mgz. I was wondering if there was a way to "extract" a new wm.mgz from an edited ribbon.mgz. My second problem involves trying to anchor a sulcus on the ventral surface during the averaging process (the mid-fusiform sulcus) in hopes of getting a more accurate representation of the sulcus on the average surface. As it stands there is a bit too much variability of the average MFS location when compared to the individual subjects' MFS location, and we'd like to reduce that if possible. Is there any way of using a label, or something similar, that can be considered as an "anchor" during the averaging process? Thanks and regards, Miggy Chuapoco 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.
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Miggy Chuapoco