[Insight-users] little double about the output of Geodesic Active Contours Segmentation with Shape Guidance

[Juan Cardelino]

I'm using that filter, and I'm experiencing an odd behavior, which
probably cames from the fact that I don't fully understand some
details of its output. Let me describe briefly the scenario.

After some trial and error, I arrived to the conclusion that the
output level set is not registered with the image to segment, but you
have the parameters of the transformation to put it in place. My
validation procedure is to rotate and translate the shape and overlay
it on top of the image.
For that purpose, I assumed that the angle given by the algorithm is
the angle of the principal axis of the shape. So, to put the shape in
place, I compute the current angle of the output shape and compute the
difference of them, which should be the rotation angle of the output
shape.
I did this in one example, and it worked, however it failed in a
second one. So, seems that the first one worked by chance.

So, here are my concrete questions:
1.- what is the exact meaning of the transform parameters given as output?
2- for instance, is the angle of the final location  (absolute) of the
shape? or is the amount  (relative) that needs to be rotated to get to
the final location?
3- in the first case, how these angles are measured of the angles?
counter-clockwise from the x-axis?


Thanks in advance.
Best regards,
                      Juan