Chapter 11:

Question 1

What is image registration?

Answer 1

Lining two images up so they look the most alike. Formally: determining a geometrical transformation

Question 2

What is an image feature?

Answer 2

Something easily recognizeable in an image.

Question 3

Name different kinds of image landmarks:

Answer 3

1) anatomical - expert assigned biologically meaningful
2) mathematical landmark - mark on curve placed through math e.g. max curvature
3) pseudo landmark - mark that is on a curve constructed based on anatomical or mathematical landmark

Question 4

What is an objecti9ve function?

Answer 4

A function that calculates the difference between two images.

Question 5

Optimization problem for registration?

Answer 5

For a given registration optimize the registration variables to minimize cost function.

Question 6

What is a rigid transformation?

Answer 6

Translation + rotation = keeps same coordinate system

Question 7

What is a similarity transform?

Answer 7

translation + rotation + isotropic scaling

Question 8

List the 4 elements of intensity based image registration:

Answer 8

1) Geometrical trnaform
2) The similarity -measures goodness of trnaform
3) OPtimization of transform parameters tooptimize similarity
4) Regularization -securing that only reasonable transformsare done