[vtkusers] Converting a point on a vtkActor to an Image coordinate

[Shaan Malik]

I've produced an actor from vtkMarchingCubes, which takes a vtkImageData 
as input.
I want to be able to select the actor in a view and convert the selected 
point on the actor to the corresponding point on the original image.

In order to get an actor position, I'm using a vtkPicker, but sometimes 
clicking not on the actor, the picker returns what it thinks is a valid 
point on the actor. At first I thought this was because the bounding box 
the Picker uses wasn't a tight fit, but actually the points it returns 
can fall outside the bounding box of the actor. This makes me assume 
vtkPicker can't be used to reliably check if a position falls on the 
actor or not?
The second stage would be to convert the actor position to a position on 
the original image, but I'm not sure how to go about doing this.

I'm trying to implement a similar feature to what occurs in ITKSNAP, 
where clicking on a segmentation changes the positions of the orthogonal 
views.

Hope someone can help,

Thanks,
Shaan