Hi Andrew --

First the good news:

Indeed, there is a much simpler (and quicker way):

mris_pmake --subj <subject> --hemi <hemi> --surface <surface> --mpmProg autodijk --mpmArgs vertexPolar:0 --mpmOverlay euclidean

If called with hemi=lh and surface=smoothwm, it will create an output overlay called lh.autodijk.crv. Each index in the overlay will have a value equal to the shortest distance along the edges of the lh.smoothwm from a start (or polar) vertex, in this case 0, to that particular vertex index in the overlay.

The whole operation (including reading all the necessary inputs) takes about 10 seconds on a current-ish Linux workstation (with bogoMips -- type "cat /proc/cpuinfo" and look for bogomips -- of about 5000)

Now the bad news:

* The version of 'mris_pmake' in FreeSurfer stable (5.1) is horribly old. And broken. For a start, notice that the all the --surface0 --curv0 --curv1 arguments are no longer required.

* The 5.1 version of mris_pmake moreover has a bug in the euclidean overlay code that means it won't work.

Better news:

* This is all fixed in 5.2 and the above command line I gave will work as advertised.

* If you can't wait for the new FS, I'm sure we can send you a compiled "preview" binary of 'mris_pmake' that will work.

* I really really need to update the documentation on mris_pmake... I'd really like to have that ready for the 5.2 release.

On Tue Jan 29 17:20:05 2013, Andrew Bock wrote:

Dear all,

I am trying to calculate the distance from a reference vertex to all other vertices in a surface mesh (or ideally to a subset of vertices), moving along the actual surface (e.g. smoothwm).

I am able to loop through the pairs of vertices in which I'm interested, using mris_pmake in each loop, but this is quite time consuming. I have tried using '--mpmProg autodijk' as described in the mris_pmake help file, but have been unsuccessful at getting this command to work.

The current code I am using for each pair is:

mris_pmake --subject <subject> --surface0 smoothwm --curv0 sulc --curv1 sulc --mpmOverlay euclidean --mpmProg pathFind --mpmArgs startVertex:###,endVertex:###

Any suggestions to cut down my processing time, while still calculating along the surface, would be greatly appreciated.

Andrew

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

-- Rudolph Pienaar, M.Eng, D.Eng / email: rudolph@nmr.mgh.harvard.edu MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging 149 (2301) 13th Street, Charlestown, MA 02129 USA