If you did followed the steps in Spherical Registration (page 46) of that tutorial, then you should have results in the ATLAS_LEFT_HEM directory.
But rather than using the colin target, use this atlas target dataset (PALS_B12.LR): http://sumsdb.wustl.edu/sums/archivelist.do?archive_id=6057499 This has the 73730 standard mesh sphere. If you register your subject to this target, then you will have your deformed surface in the extracted atlas target directory on the 73730 mesh. If you have Freesurfer surfaces, I suspect the pipeline will be the shorter path to what you want. On 04/12/2011 05:06 PM, Colin Davey wrote: > Thanks Donna! > > >> http://brainvis.wustl.edu/wiki_linked_files/documentation/Caret_5.5_Tutorial_Segment.pdf >> > > > I have gone through that tutorial, but it doesn't change the number of > vertices of the mesh (unless I missed something). As you said, "changing the > geometry doesn't change the mesh". > > I'm asking about the ability to *resample* meshes to 73,730 nodes. "Using a > strategy introduced by Saad et al (2004), the various surface > configurations...were resampled so that they were represented using the > 73,730-node 'standard-mesh'..." (A PALS Atlas of Human Cerebral Cortex, Van > Essen 2005). Was that in Caret, or was that done outside of Caret? > > I haven't gotten the Freesurfer process under control yet, and will no doubt > have questions about that before long (some basic newbie ones no doubt). > > On Apr 12, 2011, at 2:41 PM, Donna Dierker wrote: > > >>> Thanks for the reply! >>> >>> I was under the impression that inflating, flattening and registration >>> didn't change the number of verteces in the mesh. That way, one can easily >>> go from a point in the flat mesh to the corresponding point in the >>> original space. (ie, if a vertex on the flatmap is vertex number 12,329, >>> than one need only get the coordinates of the 12,329th vertex of the >>> corresponding cortical mesh.) >>> >> Correct -- changing the geometry doesn't change the mesh. >> >> >>> I couldn't find a tutorial that shows how to do the surface-based >>> registration to an atlas such that it changes the number of vertices (for >>> example to '74k'). >>> >> Look for the word "Registration" in the title, e.g.: >> >> http://brainvis.wustl.edu/wiki_linked_files/documentation/Caret_5.5_Tutorial_Segment.pdf >> >> But as I said earlier, if you use the pipeline, there is a command line >> step near the end of the postborder.sh script that runs spherical >> registration. Everything in preborder.sh and postborder.sh is leading up >> to this final step. >> >> >> >>> On Apr 9, 2011, at 8:25 PM, David Van Essen wrote: >>> >>> >>>> Colin, >>>> >>>> Segmentation produces a 'native mesh' whose exact node number cannot be >>>> prespecified in Caret (or in other segmentation algorithms I know >>>> about). >>>> >>>> The 73,730-node mesh ('74k' in our latest lingo) arises from >>>> surface-based registration to an atlas - either the macaque F99 atlas (a >>>> 74k_f99 mesh) or the human PALS-B12 atlas (a 74k_pals mesh) - using the >>>> appropriate tutorial. >>>> >>>> David VE >>>> >>>> >>>> On Apr 8, 2011, at 6:48 PM, Colin Davey wrote: >>>> >>>> >>>>> Hi, >>>>> >>>>> How do you create a 73,730 standard-mesh with Caret? I assume you begin >>>>> by going through segmentation, flattening and registration process from >>>>> the tutorial of the same name. But then what? >>>>> >>>>> Cheers, >>>>> - C >>>>> >>>>> *************************************** >>>>> Colin Davey >>>>> Scientist >>>>> Senior Software Engineer >>>>> >>>>> Electrical Geodesics, Inc. >>>>> 1600 Millrace Dr. St 307 >>>>> Eugene, OR 97403 >>>>> >>>>> Learn Boogie Woogie Piano >>>>> http://www.ColinDavey.com/BoogieWoogie >>>>> *************************************** >>>>> _______________________________________________ caret-users mailing list [email protected] http://brainvis.wustl.edu/mailman/listinfo/caret-users
