[Insight-users] mesh to volumetric image

[Yan Pingkun]

Celina & Luis,

Thanks for your reply! Now I am trying to do this work in vtk. Are you
familiar with the vtkProbeFilter? I may use planes to probe the
polydata. Hope it works.

Cheers,
Pingkun

-----Original Message-----
From: ci42 at columbia.edu [mailto:ci42 at columbia.edu]=20
Sent: Wednesday, March 17, 2004 12:37 AM
To: Yan Pingkun
Cc: Luis Ibanez; insight-users at itk.org
Subject: RE: [Insight-users] mesh to volumetric image



 Pingkun,

  I agree that a deformable mesh could be "projected back" into the
volume... I know that Ting Chen implemented it but I don't think this is
part of the itk. The problem is that the moment you grow a deformable
mesh, you don't know its location with respect to a corresponding
volume. What could be done: 1. When a mesh is generated, keep track
where its vertices are located with respect to the volume (in way way,
maintain dual mesh-digital representation of the 3D shape OR 2.
Implement an algorithm that will intersect "mesh with volume", that can
be a generic useful tool with many applications...

 none of the features is available now under the itk,

 -Celina



On Tue, 16 Mar 2004, Yan Pingkun wrote:

> Hi, Luis,
>
> Thank you for your quick reply!
>
> >Hi Pingkun,
>
> >Your 3D objects,.... are they 3D Images ?
> >or geometrical representation of objects ?
> >(like surfaces...?)
>
> My objects are in geometrical representation. They can be represented=20
> by itk::mesh or vtk::polydata.
>
> >When you use the Mesh deformable model, you
> >usually expect this deformed surface to be
> >the final result of your process.
>
> That's true!
>
> >If you want to get a deformation field that
> >you can use for wrapping an image or a set
> >of geometrical objects, you should rather
> >look at methods such as:
>
> I choose deformable model because it is a requirement in this project.

> The first step is to handle these 3D objects. An initial mesh should=20
> snap to an object surface. We just need to load one object each time.=20
> In practice, this is useless as we already know the results. So, it is

> just an experiment.
>
> After that, I'd like to apply the 3D deformable model on some real=20
> volumetric medical images just like that in the new example.
>
> Regards,
> Pingkun
>
> >   1) FEM-Based deformable registration
>
> >   2) Demons deformable registration
>
> >   3) BSplines transform for deformable
> >      registration
>
> >   4) KernelTransform based on landmarks
> >      for warping space.
>
>
>
> Regards,
>
>
>     Luis
>
>
>
>
> --------------------
> Yan Pingkun wrote:
>
> > Dear all,
> >
> > I am doing an experiment on 3D deformable models. Now I have a set=20
> > of 3D objects and I need to apply 3D deformable model filter on=20
> > them. The problem is how to convert the geometric data into=20
> > volumetric data. Any
> idea?
> >
> > Thanks!
> >
> > Regards,
> > Pingkun
>
>
>
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org=20
> http://www.itk.org/mailman/listinfo/insight-users
>