Biomedical Engineering Reference
In-Depth Information
distribution but, on closer inspection, the question occurs to one as to
why the high dose region extends so far outside the target volume.
There are two reasons for this which reinforce one another. First, as I
have already discussed
in extenso
, a safety margin has been left and,
because many of the beams are going through very complex
inhomogeneities in the base of skull, the distal margins in particular
can be quite substantial. Second, it is often the case that a neigh-
boring section has a larger target area. Because of the need to provide
safety margins in all directions, the dose needed to adequately cover
the neighboring section “spills over” into the section one is looking
at. A particular example of this arises when one is using a scattered
proton beam to treat a pear-shaped volume from the top of the pear.
In order to cover the cross section of the target volume at depth,
a larger-than-desirable field covers the smaller upper sections of
the pear. It is very important in judging dose distributions to inspect
the dose in many sections throughout the treated volume, and not just
in one or a few selected sections.
The point I am making here is a very fundamental one. Too frequently
one sees charged particle plans in which the high dose volume
tightly hugs the target volume. Such plans are testimony to a lack
of understanding of the importance of uncertainty estimation and
too much faith in the pretty colored pictures which can be painted by
the computer. I am convinced that the good local control that has
been achieved at the Massachusetts General Hospital in treating base-
of-skull sarcomas has owed a lot to the conservative treatment
margins employed.
Field patching
One not infrequently encounters the situation in which a horse-shoe
shaped target volume is wrapped around an organ to which one does
not want to deliver the full prescription dose. This problem can be
solved in one of several ways. IMPT is one, as discussed below, and
the use of a single proton field tailored distally to stay off the central
region of the horse shoe (e.g., the anterior field sketched in Figure
11.12a) is another. However, this second approach can be highly
problematic if the tissue densities proximal to the organ to be spared
are complex, and it is often avoided for that reason. The beam whose
90% isodose contour is outlined in Figure 11.12a, for example, would
never be considered in practice because of the complex inhomo-
geneities through which it passes. One could not be sure of sparing
the brain stem.
