Biomedical Engineering Reference
In-Depth Information
Unfortunately, the SOBP,
while still delivering vir-
tually no dose beyond the
high-dose region, delivers
a substantial dose proximal
~200 MeV
protons
~200 MeV
protons
target
region
target
region
to it. The entrance dose
depends on the extent in
depth of the SOBP and, to a
depth
depth
lesser extent, on its maxi-
mum penetration; typically it
can be 80% or even higher .
Figure 10.10. Schematic comparison of
high energy photon and proton beam
depth distributions. The golden area
Thus, in practice, one does
not achieve the dream of
having a beam that deposits
significant doses only in
the region of the tumor.
identifies unwanted dose delivered by
photons, but not protons. The purple
area at small depths identifies a small but
sometimes important region in which
protons lack the skin-sparing advantage
provided by high energy photons.
Nevertheless, the dose dis-
tribution of a proton SOBP is much superior to that of a photon beam
from a typical linear accelerator. Figure 10.10 illustrates the most
important ways in which the two modalities differ.
There is one further effect that bears mentioning, which occurs
equally in photon therapy, namely, the inverse square effect. Relative
to an initially parallel beam, the flux of protons, and hence the dose,
at a point a distance 'r' from the source (which may be either real, as
in scattered beams, or virtual, as in scanned beams) will be reduced
by a factor of 1/r
2
. The inverse-square effect depresses the dose at
larger depths. This is compensated for by adjusting the weights of the
upstream Bragg peaks, but this compensation unavoidably raises the
entrance dose. The practical consequence of this is that the modest
sparing of proximal tissues that a SOBP provides is further reduced,
the shorter the source-to-patient distance is. For example, for a 250
MeV beam with a SOBP width of 10 cm, the ratio of the entrance
dose to dose within the SOBP is: 64% at 30 m source-to-isocenter
distance; 77% at 3 m; and 127% at 1 m. For this reason, one tries in
proton beam therapy to keep the virtual source at least 2 m from the
isocenter and preferably 3 m or more.
For scanned beams, if the scanning is achieved by a pair of dipole
magnets offset from one another in distance along the beam axis,
there will be two virtual sources at rather different distances from the
patient. This somewhat complicates calculation of the inverse-square
