Speaker
Description
Summary
Current radiotherapy treatment planning is based in local
software tools (such
as Pinnacle, XiO, Oncentra, Corvus, etc.) running on
workstations, i.e. they run
at hospital premises. Specialized personnel compute treatment
plans either
employing their previous knowledge and experience or
class-solution based on
trial-and-error method or, for more complex treatment plans,
employing built-in
optimization tools. These software tools, called treatment
planning systems
(TPS), are subject to very severe constraints on computer power
and time to
produce results, due to hospital workload and limited access to
new algorithms.
The requirements in maximum computation time forces the TPS tools
to make
some approximations both in the dose calculation engines and
optimization
algorithms. The most accurate dose calculation techniques that
these codes
implement are based on convolution/superposition (C/S)
techniques, which
suffer from certain limitations in high density gradient regions.
Usually, the
treatment is based on the Conformal Radio Therapy (CRT) where the
incident
angles, fixed beam energy and exposure time are selected,
conforming the
beam to have the same shape of the tumour for each angle. The
more recent
Intensity Modulated Radiation Therapy (IMRT) techniques allow
selecting the
intensity for each incident angle in great detail.
State-of-the-art conformal and complex radiotherapy treatments
such as IMRT
are calculated and optimized employing this kind of tools.
Treatments are
tailored to maximize the dose to the planned target volume (PTV)
while
minimizing the dose to surrounding tissues, specially the organs
at risk (OAR),
within the limits specified by the doctors. This is the main
concept in the
definition of the objective function of the optimization which is
solved using
several well-known techniques such as simulated-annealing, linear
programming or mixed integer programming.
At the end of the project, a single portal will make available to
radiotherapists
several techniques to optimize treatments and verify them, taking
into account
at the same time the new paradigm of Service Oriented
Architectures (SOA). On
the first stage, we plan to include the next tools:
• Monte Carlo methods for treatment verification. They accurately
model the interaction of radiation with matter and are the
dominant standard in
dose calculation techniques. They achieve more accurate results than
convolution/superposition methods, at a higher computational
cost. Although
forthcoming achievements may render Monte Carlo a valid
near-real-time
treatment planning alternative, it currently plays an outstanding
role as a
validation technique. Therefore, the eIMRT platform will use it
to verify the
results of commercial treatment planning systems for any kind of
radiotherapy
plan of external photon beams.
• CRT and IMRT optimization algorithms. New web services will
implement an open access point to exhaustive yet computationally
intensive
IMRT and CRT optimization algorithms based on mixed Monte Carlo
C/S dose
computation algorithms which will produce optimized treatment
plans of a
quality (in terms of dose conformation and organ sparing) hardly
achievable
with commercial TPS.
• Finally, there is a joint effort of the international groups
involved in
this project to establish a public data repository with
anonimyzed CT scans,
treatments and other relevant information, which may be useful in
the future to
reproduce results, mine knowledge and train specialists.
Actually we have implemented the treatement verification on GRID,
making
some test using EGEE infrastructure. The verification process
comprises several
sequential steps, some of which have large computational
requirements, others
with a large number of simultaneous and short jobs which have to
be run in a
brief time period (no more than a few hours).
