[Insight-users] Re: Image or Mesh to Tubes

[Luis Ibanez]

Kevin,

As you pointed out, if you want to fit a tree-tubular structure
to the neurons in your image, doing this in a progressive manner
is the way to go.

You could start by fitting a tube to the trunk of the dendrite,
then from it start adding/fitting tube to the next level of
branching, and so on. In this way you will only be fitting one
tube at a time, and one of the tube ends will be almost anchored
in the parent of the tree structure.

I'm sure Julien and Stephen will provide much valuable input
on how to do this efficiently.



    Regards,


       Luis


---------------------
Kevin H. Hobbs wrote:
> Thank you for your comments.
> 
> On Wed, 2006-08-02 at 07:22 -0400, Luis Ibanez wrote:
> 
>>You may want to get in touch with Julien Jomier and Stephen Aylward
>>(cc in this email). They developed most of the Spatial Objects hierarchy
>>in ITK, and they have been working in using the TubeSpatialObject for
>>representing vascular trees by doing Model-to-Image registration.
>>
> 
> 
> I definitely look forward to their input.
> 
> 
>>For some background on Model-to-Image registration, you may want to
>>look at the final section of the Image Registration chapter in the
>>ITK Software Guide
> 
> 
> I'll reread that chapter. I was initially worried that adding branches
> would change the number of parameters in a model, and might really make
> optimization interesting. Now that I've slept on it however, I suppose
> there's no reason that one piece couldn't be fit at a time, and added to
> the tree.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users