Dear all,
I have a problem regarding the label volume editor. I used FSL-melodic to analyse resting state data. Now I would like to create ROI's for further analyses by using the label volume editor. I used my resting state file (melodic_IC) as the label volume and the structural image as the support volume.
However, the images do not fit together in the way that the resting state image ist shifted. The same happens when using the standard MNI template instead fo the structural. Only the funtional data fits to the melodic resting state data. I attached the images for all three cases.
Do you by any change have an idea what the problem might be?
Thank you very much in advance,
greetings
Christiane
label volume editor problem with overlay
-
ChristianeJ.
- Posts: 4
- Joined: Tue Mar 27, 2012 8:42 am
label volume editor problem with overlay
- Attachments
-
- resting state components to functional data
- melodic_IC_functional_2.jpg (123.02 KiB) Viewed 3533 times
-
- resting state components to standard MNI template
- melodic_IC_standard.jpg (139.85 KiB) Viewed 3533 times
-
- resting state components to T1 image
- melodic_IC_T1.jpg (250.27 KiB) Viewed 3533 times
-
riviere
- Site Admin
- Posts: 1361
- Joined: Tue Jan 06, 2004 12:21 pm
- Location: CEA NeuroSpin, Saint Aubin, France
- Contact:
Re: label volume editor problem with overlay
Hi Christiane,
It's probably a matter of applying the coordinates transfomations. Anatomist, by default, does not use transformations information in volumes headers, partly for obsolete historical reasons, and partly because it would force it to make assumptions about where such transformations are going (are the destination referentials in several volumes the same or not). But you can make Anatomist use them: in the main window, select the volume(s) and right-click on them. In the menu, select "referential/load information from file header". If you're lucky, it will be enough. But it depends on which information is actually present in images headers.
If the resting state data is intra-subject and has been acquired together with the T1, then there is probably a fixed translation between both, and each image has a transformation to the scanner space. But, especially if using NIFTI format, ther is nothing saying that both images are from the _same_ scanner, and thus that transform go to the same scanner-based referential. So this has to be set manually in Anatomist referentials/transformations window. Or there is a settings option to make it assume it, but it's disabled by default.
If data has been normalized into the MNI space, then the transformation from each volume to the MNI space must have been set in input volumes (SPM does so, and I think FSL also). But, here again, something has to tell that the transformations are going to the same normalized space. There is a code in NIFTI format to say so, but I know SPM for instance doesn't set it when normalizing images, and I don't remember for FSL. So here again, it it doesn't work at once, you have to tell Anatomist that both destination referentials are the same.
You can look at pieces of doc on how to use referentials and transformations here in the tutorial, or in the manual.
Denis
It's probably a matter of applying the coordinates transfomations. Anatomist, by default, does not use transformations information in volumes headers, partly for obsolete historical reasons, and partly because it would force it to make assumptions about where such transformations are going (are the destination referentials in several volumes the same or not). But you can make Anatomist use them: in the main window, select the volume(s) and right-click on them. In the menu, select "referential/load information from file header". If you're lucky, it will be enough. But it depends on which information is actually present in images headers.
If the resting state data is intra-subject and has been acquired together with the T1, then there is probably a fixed translation between both, and each image has a transformation to the scanner space. But, especially if using NIFTI format, ther is nothing saying that both images are from the _same_ scanner, and thus that transform go to the same scanner-based referential. So this has to be set manually in Anatomist referentials/transformations window. Or there is a settings option to make it assume it, but it's disabled by default.
If data has been normalized into the MNI space, then the transformation from each volume to the MNI space must have been set in input volumes (SPM does so, and I think FSL also). But, here again, something has to tell that the transformations are going to the same normalized space. There is a code in NIFTI format to say so, but I know SPM for instance doesn't set it when normalizing images, and I don't remember for FSL. So here again, it it doesn't work at once, you have to tell Anatomist that both destination referentials are the same.
You can look at pieces of doc on how to use referentials and transformations here in the tutorial, or in the manual.
Denis