Medical Image Registration: Theory, Algorithm, and Case Studies in Surgical Simulation, Chest Cancer, and Multiple Sclerosis - Biomedical Image Analysis - page 20

Population

Displacements

Rotations

∆ y ∆ z γ

∆ x α

β

8b 8b 6b

8b 6b 6b

(a)

Displacements Rotation

∆ x ∆ y θ

8b 8b 6b

(b)

Figure 1.4:

Different gene structures.

1 5

Generations 2 50

gene 1 gene 1 gene 1

gene 1

fit. 5%

fit. 7% fit. 25%

fit. 95%

gene 2

gene 2 gene 2

gene 2

.
.
.

.
.
.

.
.
.

fit. 1% fit. 10% fit. 60%

fit. 97%

gene 25 gene 25

gene 25

gene 25

.
.
.

.
.
.

.
.
.

.
.
.

fit. 2%

fit. 9% fit. 98%

fit. 99%

gene 100

gene 100 gene 100

gene 100

fit. 7% fit. 0.5% fit. 40%

fit. 95%

Figure 1.5: An illustration of the GA use in calculating the transformation pa-
rameters. The process starts by coding the transformation parameters into one
string. An initial population of strings is randomly generated. Applying the se-
lection, crossover and the mutation operations, new generations are obtained.
Notice how the registration fitness function average increases with the genera-
tions. Maximum fitness can be reached eventually.