Biomedical Engineering Reference
In-Depth Information
Apertures
The design of apertures has already been discussed in Chapter 10. It
is largely a matter of trivial geometry. For scattered broad beams
there are only two wrinkles: there can be aperture-edge effects which
give rise to mainly superficial dose perturbations; and neutron
production in apertures and/or blocks, while quite small, is not
entirely negligible. Both have been discussed in Chapter 10.
Scanned beams do not have either disadvantage. The irradiated
volume is entirely defined by the pattern of pencil beams which are
applied and there is no material to produce either edge-effects or
neutrons. One might think that one would “turn on” pencil beams
which are headed towards the target volume, and turn all others “off.”
One must remember however that scattering effects will deplete the
dose of the pencil beams lying at or near the geometric edge of the
beam, and so one must add pencil beams around the periphery of
the beam in order to ensure that there is no dose deficit to the target
volume at its edges. This margin is needed, just as for photons, even
if the target volume is the planning target volume (PTV), in order to
compensate for dose fall-off in the penumbral region of the beams.
Scanned beams do, however, have one characteristic which must be
taken into account. In most implementations, there are two sweeping
magnets, spatially offset from one another as shown in Figure 10.21
of Chapter 10. This means that there are two spatially offset virtual
sources of the beam - one for each direction of scanning. Thus, the
beam's-eye view is more complicated than a simple perspective
projection, complicating both the computation of BEV images in the
treatment planning system and the design of the aperture.
Compensators
A compensator is a device, be it real or virtual, for making range
modifications in the beam that reaches the patient so that wherever a
greater beam penetration is required in the patient the compensation is
less, and where less penetration is desired, the compensation is more.
Figure 11.7a shows schematically the consequence of making no
compensation for the bone sliver - one would have a cold region in
the target volume. Figure 11.7b shows the consequence of making an
“exact” compensation - that is, when the compensator is modified
only where it geometrically shadows the inhomogeneity. The dose
distribution in Figure 11.7b looks satisfactory (although it ignores
scattering effects which would, in fact, prevent perfect compen-
sation). However, if there were a small mis-registration between the
