Information Technology Reference
In-Depth Information
have been used. Although these algorithms are suitable for optimization, they still
require a significant amount of time to generate the most optimal treatment plan. In
the medical field, quick intervention can be the key to the improvement of cancer
control rates. Thus, shortening the time spent generating a plan can allow the pa-
tient to be treated sooner.
As a result of reviewing the remarkable results demonstrated in other scientific
fields [4-9], Harmony Search was chosen to tackle the problem of optimizing HDR
brachytherapy for prostate cancer. Radiation dose-based optimization constraints were
evaluated in order to determine the most optimal treatment plan. Based on the results
of this investigation, Harmony Search is considered a natural fit for therapeutic medi-
cal physics, which can be applied to similar scenarios within the field.
2 Radiation Treatment Planning
The typical process of delivering radiation to a patient involves acquiring a set of im-
ages of the patient, such as a computed tomography (CT) scan. Prior to the scan, the
patient is placed in the position that they are to be treated in so that the images reflect
reproducible anatomy and geometry. Next, the scan is imported into the radiation
treatment planning system. This sophisticated system is then able to reconstruct the
patient in 3D, allow the radiation target and critical structures to be delineated, and to
place, calculate and evaluate radiation dosimetry.
The radiation oncologist will first decide what part of the anatomy will be treated
and contour those structures on the image scan. Additionally, the physician will spec-
ify the prescription dose the target must receive. After this is completed, the organs-
at-risk (OARs) are outlined. The next step is to optimize the parameters which control
the radiation, virtually simulate the radiation interactions in the patient and determine
the final planned dose to the target and the OARs. The plan can then be evaluated by
using dose-volume histograms.
2.1 Overview of Dose-Volume Histograms
The dose volume histogram (DVH) was described by Chen et al. in 1987 [10]. It is a
version of the standard histogram that allows the analysis of a tabulated data set, with
regards to the frequency each data point occurs. The data is separated into bins or
ranges, and each time a data point falls into a bin, the frequency of that bin is in-
creased by one. The frequency of data is plotted against the bin value.
In the dose volume histogram, dose values (typically measured in cGy or Gy) are
tabulated either by points or voxels within a specific organ structure. Each structure is
calculated separately and the plot is shown as the volume of the structure receiving a
given dose or higher as a function of dose. This form of the DVH is known as a cu-
mulative DVH, since the plot shows the volume that receives a specific dose or
higher. The bin value can be set to any desired dose interval, which will modify the
coarseness of the DVH. An example of a DVH with a bin value of cGy is given in
Figure 1.
Search WWH ::
Custom Search