[Insight-developers] contribution of JHU Computational Anatomy algorithms based on Large Deformation Diffeomorphic Mapping and Laplace-Beltrami eigenanalysis

[Michel Audette]

Hi Arnaud,

thanks for your kind response. I'll have a look at the IJ paper and discuss
it with my collaborators at JHU.

Dear fellow developers,
on another note, I should correct my previous post, based on correspondence
from Michael Miller. The contribution by JHU to the IJ and, hopefully ITK,
will be the Laplace-Beltrami eigenanalysis, but not the LDDMM. Perhaps the
community at large can revisit the latter problem once the Laplace-Beltrami
code is in place.

I apologize for the misinformation on my part.

Best wishes,

Michel


On Mon, Oct 26, 2009 at 4:44 PM, Arnaud Gelas
<arnaud_gelas at hms.harvard.edu>wrote:

> Hi Michel,
>
> Great! That's really interesting work!!!
> I am pretty sure that many persons will be interested in such application!
>
> For the Laplace beltrami laplace eigenanalysis, computing the matrix is
> "almost" done in itk.
> You can have a look at the following insight journal paper:
> http://insight-journal.org/browse/publication/202 (note that the
> corresponding class are in Review directory of itk).
>
> Now regarding the eigenanalysis, this is really computationally expensive.
> Generally people use ARPACK, but I suggest to have a look at bruno levy's
> work on the subject where he also proposed some optimization using TAUCS and
> SUPERLU.
>
> http://alice.loria.fr/index.php/publications.html?redirect=0&Paper=spectral_course@2009
>
> HTH,
> Arnaud
>
> On Oct 26, 2009, at 6:48 PM, Michel Audette wrote:
>
>  determinants
>>
>
>


-- 
Michel Audette, Ph.D.
R & D Engineer,
Kitware Inc.,
Chapel Hill, N.C.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20091027/4e5ca77d/attachment.htm>