|
|
|
Clinical background
Radiotherapy planners face the difficult task of matching two generically
conflicting planning goals: a certain curative dose must be applied to the
tumor, the target volume, in order to guarantee the treatment success with
a sufficiently high probability, while simultaneously keeping the dose
deposit in the adjacent healthy tissue, especially in the organs at risk, on
an ideally low level in order to avoid fatal complications.
The main topic of the computer supported serach for patient specific
optimal therapies is the definition of an adequate plan evalution method.
Current treatment planning software evaluates plans with single values,
that result by means of an adequate weighing from the individual evaluations
of the dose distributions in all relevant entities, i.e. the target volume and the
organs at risk. This approach does not satisfactorily represent the clinical
practitioner's organ specific judgement of a treatment plan. Necessary
modifications of the artificial weights cannot be realized dynamically, thus
leading to a time consuming trial-and-error search process.
Project goals
With the work done in several projects, Fraunhofer ITWM contributes with
different departments being involved in cooperation with its partners from
the fields of medicine, physics, computer science and mathematics, to an
improved treatment planning in the radiotherapy by means of open fields
and the intensity-modulated radiotherapy (IMRT).
- Main field of activity of the projects is the:
 Optimization of the therapy setup
in the department
Optimization of Fraunhofer ITWM.
Based on a model of
treatment planning by means of multicriteria optimization, physical therapy
setups are computed. Instead of the single valued plan evaluation, multiple
evaluations of the dose distribution in all involved entitities occur, which enter
the optmization separately. Based on the Pareto solution concept for
multicriteria problems, representative systems of efficient solutions are
computed and stored in a database, which the oncologist can search
interactively in order to find the treatment plan fitting best to the specific
clinical case. The navigation tool developed and patented by Fraunhofer ITWM
visualizes the current treatment plan proposal and the possibilities in the specific
clinical case in an intuitive way.
-
An important basis of the optimization modul is an efficient and exact calculation
of the dose distribution, which is based on a
Modelling of the dose absorption by means of transport equations.
New approaches of the department
Transport processes of Fraunhofer ITWM
aim for a consideration of the classical pencil-beam models as special
discretizations of physical transport equations. Project goal is a transfer of
strategies, that proved successful for the solution of transport equations in other
projects, to the dose computation in radiotherapy planning.
-
A major problem in clinical radiotherapy practise is the patient geometry, that
continually changes during the 30 days lasting treatment phase. Treatment
parameters computed at the beginning, have to be fitted during the treatment to
the current patient geometry. This requires an
adaptive matching of the therapy plans
during the whole treatment phase. A first step of this adaptation process, the
automatic detection of changes between two images of the patient geometry is
modelled by the dapartment
Models and Algorithms in Image Processing of
Fraunhofer ITWM. One applies methods from elastic registration, that can
derive the transformation merely from the image data. For an increase of
efficiency, a multiscale approach is used - the transformation map.
-
In the field of intesity-modulated radiotherapy planning (IMRT), the efficient
physical realization of the therapy plans by means of multileaf collimators
is relevant of clinical acceptance. The department
Optimization of Fraunhofer
ITWM models and solves the optimal setup for the realization by means of static
step-and-shoot methods as a network-flow problem in the field of discret
optimization.
Stand: 05.08.03 |
