Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Hi Margaret,
How does the talairach transform look?
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The one using the orig is really horrible looking, very distorted and rotated. I tried using the brain volume, and it looks better, but also has a rimming/shadowing effect.
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi Margaret,
How does the talairach transform look?
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi again,
If the shadowing/rimming effect appears to be because of an artifact in one of the scans, then you might be better off to exclude that acquisition. But if each of the individual scans looks okay, then it does sound like improving your lousy average is dependent upon getting a better talairach transform. You can do this by trying various talairach scripts or even registering it manually.
On Wed, 2 Feb 2005, Margaret Duff wrote:
The one using the orig is really horrible looking, very distorted and rotated. I tried using the brain volume, and it looks better, but also has a rimming/shadowing effect.
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi Margaret,
How does the talairach transform look?
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
i would not advise averaging data from two separate visits. use one or the other.
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi again,
If the shadowing/rimming effect appears to be because of an artifact in one of the scans, then you might be better off to exclude that acquisition. But if each of the individual scans looks okay, then it does sound like improving your lousy average is dependent upon getting a better talairach transform. You can do this by trying various talairach scripts or even registering it manually.
On Wed, 2 Feb 2005, Margaret Duff wrote:
The one using the orig is really horrible looking, very distorted and rotated. I tried using the brain volume, and it looks better, but also has a rimming/shadowing effect.
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi Margaret,
How does the talairach transform look?
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
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
On Wed, 2 Feb 2005, Brian T. Quinn wrote:
i would not advise averaging data from two separate visits. use one or the other.
The auto-align feature aims to offer a fair degree of consistency between session scans, but if this was not the case (and it often isn't, because you cannot control for the position of the coil, etc.), then definitely be prepared to ditch the worse scan. Unfortunately!
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi again,
If the shadowing/rimming effect appears to be because of an artifact in one of the scans, then you might be better off to exclude that acquisition. But if each of the individual scans looks okay, then it does sound like improving your lousy average is dependent upon getting a better talairach transform. You can do this by trying various talairach scripts or even registering it manually.
On Wed, 2 Feb 2005, Margaret Duff wrote:
The one using the orig is really horrible looking, very distorted and rotated. I tried using the brain volume, and it looks better, but also has a rimming/shadowing effect.
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi Margaret,
How does the talairach transform look?
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
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
Before you give up you could try running distortion correction on each of those volumes before you try averaging/motion correction again...
On Wed, 2 Feb 2005, Margaret Duff wrote:
The one using the orig is really horrible looking, very distorted and rotated. I tried using the brain volume, and it looks better, but also has a rimming/shadowing effect.
On Wed, 2 Feb 2005, Evelina Busa wrote:
Hi Margaret,
How does the talairach transform look?
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Margaret,
what are you using to align them? If they are from different sites, then they may not be possible to really align, due to differences in shims, read-out directions and bandwidth. Reconning with a single image is certainly doable (we've done hundreds), but is more labor intensive.
cheers, Bruce
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi, Both images are from here on the old bay 2 scanner, they just occurred 1 month apart. Could you describe a little for me how to use mri_motion_correct and how it differs from setup structural scans in csurf? I checked the wiki, but it did not have tons of info. Also, what are the major differences in using only one scan as opposed to two for the reconstruction? Thanks for your help, Margaret
On Wed, 2 Feb 2005, Bruce Fischl wrote:
Hi Margaret,
what are you using to align them? If they are from different sites, then they may not be possible to really align, due to differences in shims, read-out directions and bandwidth. Reconning with a single image is certainly doable (we've done hundreds), but is more labor intensive.
cheers, Bruce
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The best thing is to follow the instructions in recon-all -help. recon-all will automatically use mri_motion_correct2. Basically, you just need to put your two scans in mri/orig/001 and mri/orig/002. However, if there is an artifact (as Ev suggests), then you'll have to use just one of the scans.
doug
Margaret Duff wrote:
Hi, Both images are from here on the old bay 2 scanner, they just occurred 1 month apart. Could you describe a little for me how to use mri_motion_correct and how it differs from setup structural scans in csurf? I checked the wiki, but it did not have tons of info. Also, what are the major differences in using only one scan as opposed to two for the reconstruction? Thanks for your help, Margaret
On Wed, 2 Feb 2005, Bruce Fischl wrote:
Hi Margaret,
what are you using to align them? If they are from different sites, then they may not be possible to really align, due to differences in shims, read-out directions and bandwidth. Reconning with a single image is certainly doable (we've done hundreds), but is more labor intensive.
cheers, Bruce
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
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
Hi Margaret,
mri_motion_correct calls the mni rigid body alignment code. Use it as:
mri_motion_correct <output> <in 1> <in 2> ...
cheers, Bruce
p.s. you still are probably better off just using one volume. There's no difference in processing, but more manual intervention will probably be needed.
On Thu, 3 Feb 2005, Margaret Duff wrote:
Hi, Both images are from here on the old bay 2 scanner, they just occurred 1 month apart. Could you describe a little for me how to use mri_motion_correct and how it differs from setup structural scans in csurf? I checked the wiki, but it did not have tons of info. Also, what are the major differences in using only one scan as opposed to two for the reconstruction? Thanks for your help, Margaret
On Wed, 2 Feb 2005, Bruce Fischl wrote:
Hi Margaret,
what are you using to align them? If they are from different sites, then they may not be possible to really align, due to differences in shims, read-out directions and bandwidth. Reconning with a single image is certainly doable (we've done hundreds), but is more labor intensive.
cheers, Bruce
On Wed, 2 Feb 2005, Margaret Duff wrote:
Hello, I am trying to reconstruct brains of patients who had one structural scan during each of 2 visits, a month apart. I averaged the two images together, but since they are from different visits, they look very badly lined up. It looks like there is one image and another image overlaid and shifted by a cm or so. These are clearly wrong and will not work but I was wondering if it was OK to run freesurfer with only one of the images, or if there was a way to line these two images up better so they can be combined? Thanks, Margaret
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Brian T. Quinn -
Bruce Fischl -
Doug Greve -
Evelina Busa -
Margaret Duff