Information Technology Reference
In-Depth Information
Harmony Search in Therapeutic Medical Physics
Aditya Panchal
Department of Medical Physics, Advocate Lutheran General Hospital,
1700 Luther Lane, Park Ridge, IL 60068, USA
apanchal@bastula.org
Abstract.
Medical physics is a branch of physics that concerns the application of radiation for
therapeutic and diagnostic use in medicine. In therapeutic medical physics, ionizing radiation is
used to treat patients afflicted with cancer. The main goal in radiotherapy is to deliver a high
amount of radiation dose to cancer cells while sparing surrounding organs. The process begins
with radiation treatment planning that requires optimization of radioisotope placement or radia-
tion beam intensities. Prior optimization algorithms, such as genetic algorithm and simulated
annealing, have been widely discussed in medical physics literature. However, Harmony Search
has shown to be a superior optimization algorithm based on the results in other scientific fields.
Thus, this study investigated high dose-rate prostate brachytherapy optimization using Har-
mony Search. Results showed that the algorithm is significantly faster than genetic algorithm
and that rapid planning allows improved patient care in medical physics.
Keywords:
Medical Physics, Radiation Oncology, Brachytherapy, Optimization, Harmony Search.
1 Introduction
Radiation therapy has been used to treat cancer for over a hundred years. In that time,
significant advances have been made in both the localization of cancer and the deliv-
ery of radiation. Throughout the last decade, more precise targeting of tumors due to
new treatment modalities have allowed customization of the radiation dose which in
turn minimizes dose to adjacent critical structures and ensures that the prescribed dose
is delivered to the tumor. Some of these modalities include Intensity Modulated
Radiation Therapy (IMRT) and high dose-rate (HDR) brachytherapy.
The purpose of therapeutic radiation stems from the radiobiological effect of ioniz-
ing radiation interacting with the DNA of cancer cells. This renders them frozen in a
specific stage of the cell cycle, unable to reproduce. Since the radiation dose prescrip-
tion can be quite high, it is usually necessary to divide the treatment into multiple
fractions. These treatments are then performed until the total dose is delivered. This
enables healthy normal tissue to regenerate between fractions and also maximizes the
damage to tumor cells.
In order to deliver the radiation to a patient, the interaction of the particles must
be modeled on a radiation treatment planning system. Since these new modalities
allow significant choice in how to deliver the treatment, optimization of the radia-
tion beam intensity patterns (in IMRT) or radioactive source placement (in HDR
brachytherapy) must be performed to generate an effective plan for each individual
patient. In the past, optimization algorithms, such as evolutionary algorithms (in-
cluding genetic algorithm) [1], simulated annealing [2] and linear programming [3],
Search WWH ::
Custom Search