Stochastic method
<
f
>=
p
s
f
(
d
s
(
x
))
s
∑
d
i
(
x
)
=
D
ij
x
j
j
∑
Normal Optimization
Dose in voxel i is a sum of
the doses coming out of
each bixel, weighed with the
constraints
f
(
d
(
x
))
Is our IMRT objective
function which has to be
minimized
Robust Optimization
The objective function is
now minimized over all
possible scenarios, weighed
with the probability function
d
s
i
(
x
)
=
D
s
ij
x
j
j
∑
Dose is calculate for all
different scenarios = e.g. all
possible positions