Biomedical Engineering Reference
In-Depth Information
Weight = 1024
3
2.5
Weight = 1
2
1.5
1
Weight = 1024
0.5
Weight = 1024
50
1
2
3
Pareto front individuals
Neck (right)
Neck (left)
Neck (left and right)
FIGURE 7.6 (See color insert.) Pareto front of optimized antenna settings for hyperthermia treatment of the head-and-neck area. Each point
corresponds to an optimized antenna setting, and which one is selected is decided based on the weighting of the different goals (heat the tumor,
avoid exposing the left side of the neck, avoid exposing the right side of the neck). Changing the weighting corresponds to gliding along the Pareto
front and translates into shifting the energy deposition.
controllers that use feedback from probes or MRI thermometry,
and sometimes information from an internal simulation model,
to regulate the antenna steering over time, and optimize a ther-
mal dose goal or a target distribution while respecting constraints
or reducing treatment time. These approaches are, however, cur-
rently limited to 1D models. Time modulated hyperthermia can
be used to achieve better coverage of large (or multiple) tumors
and to obtain a reduced exposure time for individual hot spots
[130]. [130] uses a multi-goal optimization approach to obtain the
Pareto front of optimal solutions (see Figure 7.6), thus providing
the physician with a selection of optimized treatment settings and
an interactive way of selecting the most suitable one (taking into
account patient and measurement probe feedback).
coverage) to CEM43 (cumulative equivalent minutes at 43°C, a
dose specifying how many minutes of constant heating at 43°C
would be needed to achieve the same effect). CEM43 is based on
the observation that every additional degree above 43°C doubles
the cell killing rate, while every degree below 43°C reduces it by
a factor of four. This type of Arrhenius-like behavior lies at the
heart of the tissue damage approach, which directly calculates
the achieved biological effect using a kinetics-based Arrhenius
model. A dose concept based on nonequilibrium theory and
extending the CEM43 approach, while suggesting an explana-
tion for the observed temperature dependency, has been pro-
posed in [163].
Clinical and experimental data has shown dose-effect (cure,
coagulation volumes, etc.) relationships [16, 64, 154]. Quantities
specifying how much the tumor region with the poorest heating
was exposed (CEM43T90, TDmin) seem to correlate best with
treatment outcome. For high temperature treatments (refer to
Section 7.10), the evolution of necrosis and its impact on per-
fusion has been studied extensively in [9, 24, 159]. It has been
found that temperature isotherms are bad predictors for lesion
size, especially for short- and long-duration treatments [24, 74],
and that it is important to consider transient effects in the dose/
damage calculation [122].
7.7 Biological Effect Determination
The administered heat results in biological effects, such as coag-
ulation, necrosis, improved survival rates, etc. To quantify this,
two approaches are commonly used: dose concepts and tissue
damage calculations. They have been reviewed in [51, 76, 103],
and some biological and chemical background is provided in [8].
Various dose concepts exist, ranging from T90 (the 90th percen-
tile of the tumor temperature distribution, a measure for tumor
Search WWH ::




Custom Search