Biomedical Engineering Reference
In-Depth Information
registration for the data shown in Fig. 1.16. The registration process was able to
show that new parts of abnormality appeared in later scans between the period
June 2001 to June 2002.
Questions
1.
Based on the classification defined in this chapter, define all the possible
classifications of the following registration techniques:
(a) Registration by maximization of Mutual Information.
(b) Surface Signature registration.
(c) Grid Closest Point registration.
2.
Using the formulation for the Surface Signature images, derive and draw
the signature image for the following parametric shapes. (choose points
of interest on each shape)
(a) a sphere of radius r
⎡
⎤
r
cos(2
π
u
)
r
sin(2
π
u
)cos(2
πv
)
r
sin(2
π
u
)sin(2
πv
)
⎣
⎦
Sph
(
u
,v
)
=
(1.21)
(b) a cylinder of radius r and height h
⎡
⎤
r
cos(2
π
u
)
r
sin(2
π
u
)
v/
h
⎣
⎦
Cyl
(
u
,v
)
=
(1.22)
3.
Describe how the Genetic Algorithm technique can be used as an optimizer
for the MI registration technique.
4.
Derive Eq. (1.19) from Eq. (1.20).
