General Registration (BRAINS) descriptors or features used to make Affine Registration (12 DOF)

Operating system: Linux
Slicer version: 4.5.0

Recently, I used 3D slicer to make co-registration between breast tumor images (MRI). I used the General Registration (BRAINS), in which I used the linear transform initialized with “useMomentsAlign”. Then ,I used the flowing registration phases in order : Rigid(6 ODF), Rigid+scale(7 ODF) and Affine(12 DOF). As results, I have a good registration with the transformation file (.h5) containing the 12 transformations. My question is about the descriptors/features used in this registration, I means what is the operations used to found the 12 unknown variables (12 DOF). I imagine that a comparison between the two inputs images descriptors are made but which features exactly? (is it the shape, intensities …?)

(please only post once – the first post must be approved, and usually someone will see it within 1-2 hours)

1 Like

The module documentation page has lots of information on that:

1 Like