I am looking for guidance to calculate the MI between two unregistered images prior to any transformations with dipy. I have looked through documentation but have yet to figure out what to input for the transform argument for MutualInformation.setup.
Thanks for your question! I believe that one input here could be an instance of the AffineTransform class. You can initialize this with an identity transform by giving it an np.eye(3) (if your images are 2D) or np.eye(4) (if your images are 3D).
I haven’t tried this myself, though. So I am not 100% sure this would work. Let me know if you run into issues.