Biomedical Engineering Reference
In-Depth Information
Questions
1.
What is image registration?
2.
Why is the algorithm-based registration still required given the success of
hybrid imaging devices?
3.
What is cross-entropy? What is reversed cross-entropy? What is symmetric
divergence?
4.
Why the cross-entropy is a generalized similarity measure for image reg-
istration?
5.
When is cross-entropy (reversed cross-entropy, symmetric divergence)
maximized? When is it minimized?
6.
How can one get a favorable priori pdf?
7.
What are common unfavorable priori pdf's?
8.
What are the general steps for a similarity-based registration?
9.
Why does cross-entropy, reversed cross-entropy, and symmetric diver-
gence based registration suffer from numerical instability? How to deal
with it?
10.
What are the requirements for a good similarity measure?