460 / 1224
The method …is ready!
(and it’s convolution/superposition)
•
I can calculate D
1sc
@ any point from any
point, e.g.:
i
θ
r
j
-r
i
j
dVθ)
,ir-jrh(
1sc,i
S
jρ
i
ρ
θ),ir-jrh( dV iρ
1sc,i
S
jρ
1
jh 1sc,i
R
jρ
1
=
1scD
=
=
=
→
j
i
and the total dose to j would be:
V
dV
jh
1sc,i
S
jρ
iρ
=
1scD
j
BUT…:
1. I need to be efficient
(reduce the # of the N
6
evaluations required)
& work
with finite voxels
2. I need to account for inhomogeneities
3. I need to account for higher order of scatter (D
2sc
, D
3sc
, …, D
msc
)
4. I need to work with real sources
5. I need to account for finite patient dimensions