Well that certainly is enormously helpful, thanks Matt. There's certainly lots of interesting looking commands in wb now. That bingham estimation thing is especially intriguing.
On a separate note: I've got a huge label.gii file. It's from a caret surface registration. I want to extract just one paint from one column and save that as a label.gii and use it as a seed for probtrack. But it's not easy. Surely it is easy and I'm missing something? I do have old scripts that take a coord,topo and paint and output an ascii surface for this purpose, and I expect that system will still pretty much work. Probably. There's lots of ways to do it. matlab. But no fsl or wb or whatever? or mris_something? nothing on web. These days It's an obvious thing to want to do so I'm confused as to why it isn't obvious. Probably it is obvious. cheers, Colin > > ---------- Forwarded message ---------- > From: Matt Glasser <[email protected]> > To: "Caret, SureFit, and SuMS software users" < > [email protected]> > Cc: > Date: Sun, 26 May 2013 00:16:24 -0500 > Subject: Re: [caret-users] painting > Connectome Workbench can write a label table into a NIFTI volume via a > header extension: wb_command -volume-label-import > > Peace, > > Matt. > > On 5/25/13 4:38 PM, "[email protected]" <[email protected]> > wrote: > > >This is trickier, because there is no standard file format that specifies > >label mappings for volumes, as far as I know. (Please inform me if there > >is!) There is a WashU way to specify it in ifh files, so if you have a > >paint mapping, you can use Attributes: Paint and project your surface > >paint to a volume. Then save the resulting paint volume in WashU 4dfp.ifh > >format. > > > >Then you can see how it maps it in the ifh header, which is ASCII. If > >memory serves, the indices are two off from what they really are in the > >volume (e.g., voxels it says are 2 are really 0). > > > >Then you can open *your* label volume as a volume anatomy; save it in ifh > >format; and edit the ifh to add the index:paint-name mappings. Increment > >the actual intensities by 2, if that's what the format wants. Once you've > >added the label:name mappings, open the volume as a Volume Paint file. > > > >Your area color file needs to map paint names to RGB -- not paint indices. > > If there aren't too many of them, you can do this in Caret. > > > >You can map the volume as paint without going through all this, but it > >will name your labels unnamed1, unnamed2, etc. as far as I know. > > > >I hope someone knows an easier way. > > > > > >> Hi - > >> > >> I have: > >> > >> 1) a file that maps integers to RGB triplets > >> 2) a file that maps the same integers to names > >> 3) a volume of unsigned char format, in which each value is one of those > >> integers (ie a paint volume) > >> > >> > >> I want to map that paint volume to the surface, and have the resulting > >> paint file reflect the RGB triplets and the region names. the integer to > >> RGB to name relationship is constant. > >> > >> but, the interplay between paint files and area color files has always > >> confused me a bit so before I write any scripts I want to be sure I get > >>it > >> right. > >> > >> best, > >> > >> Colin > >> _______________________________________________ > >> 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 > > > > > _______________________________________________ > 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
