I am using the Similarity3DTransform in SimpleITK. I have a doubt. After I get the transform, the parameters concerning the shift are with the dimensions of the .mhd file or the new ones affected by the scaling of the transform?
Scaling/rotation/shear are usually applied before translation.
But, for computation of the number of pixels to shift must I divide the translation parameters by “Dimensions” or “Dimensions*Scaling”?
Probably just “Dimensions”, but it is usually easier to try then to figure it out from the documentation. And you might not be thinking in terms of physical space.