On 10/06/2010 10:27 AM, Goulas Alexandros (PSYCHOLOGY) wrote: > > Dear Caret users, > > > > I am trying to visualize the result of a parcellation scheme. I would > like to ask how a paint file can be generated. I would like to create > something like the paint file for the different sulci as it is shown > in the start up guide. The way I do the visualization now is making a > niftii file where each voxel has a unique integer that codes for the > module that it has been assigned to. When I import it to the Caret > (map volume to surfaces option) by choosing the appropriate color > mode, the different regions appear with different colors. > You want to map this as a paint/ROI volume, rather than as a functional/metric volume. > > And this bring us to the next question: In the map volumes to surfaces > menu, there are the options to choose the appropriate Caret surface. > Despite the fact that I have inflated , flat and very inflated > surfaces (as .coord files) only the fiducial is offered as an option. > This can cause a problem since when I import the volume some regions > might be colored just because their tip is near by a gyrus within my > ROI for instance (neighbouring gyri will have few scattered voxels > colored if eg the interpolation method is used ). > Never use interpolated voxel method with paint/ROI volumes. Use enclosing voxel. It is like the difference between trilinear and nearest neighbor. In this context, you need to preserve the intensities -- not weight them with nearby voxels' intensities. > > Is there a way to explicitly map the volume to a specific surface (eg. > the very inflate one)? > This almost never makes sense. The volume is almost always aligned to some fiducial surface, whether it be a white matter, midthickness, or pial representation. Unless the volume was created by projecting a ribbon around the inflated or very inflated surface, then you wouldn't have a volume aligned to an inflated surface. It would not map properly.
The way to avoid the problem above is to use the paint/ROI volume option and use enclosing voxel method. > > Some help with these issues would be great. > > > > Many thanks in advance and kind regards, > > > > Alex > > ------------------------------------------------------------------------ > > _______________________________________________ > caret-users mailing list > [email protected] > http://brainvis.wustl.edu/mailman/listinfo/caret-users > _______________________________________________ caret-users mailing list [email protected] http://brainvis.wustl.edu/mailman/listinfo/caret-users
