|dc.description.abstract||This study deals with the characteristics, dosimetry and
applications of linear accelerators. Firstly the theory
and operation of a linear accelerator is discussed, for
a detailed knowledge of its operation is essential for
an understanding of the parameters which may influence
the electron and X-ray beam characteristics.
The modern linear accelerator uses a specially designed
waveguide. Electrons generated by an electron gun, are
accelerated by means of a pulsed radio frequency electromagnetic wave. This RF-wave of the TM 01 mode may be generated by various devices like magnetrons and klystrons.
The frequency of the R.F. generator has to be regulated to
within close limits of the tuned waveguide frequency to ensure that the electrons always have the same energy.
Another important component of a linear accelerator is
the bending magnet. The most common type of magnet in use
today is a 270° bending magnet which seems to be superior
to the 900bending magnet. In the case of the 90° magnet,
the beam deviation due to small variations in magnet current, is much more pronounced than in the case of the 270°
Further, the linear accelerator produces electrons and
X-rays at energies which are much higher than the y-rays of
Cobalt units. Therefore the main essentials of the
theory of X-ray and electron interaction with matter,
as well as dosimetrie techniques are described.
The results obtained for the Mevatron 6 and Mevatron 8
linear accelerators in present use at the Institute of
Radiation and Isotopes, Bloemfontein are presented. A
very important observation was made regarding the determination of electron energies. The size of the dose chamber is extremely important and the diameter of the chamber has to be as small as possible. The most suitable
instrument for this purpose was found to be an extrapolation chamber, such as the product manufactured by S.H.M.
Nuclear. To comply with the international definition of
the roentgen and rad, the Baldwin Farmer 0,6 cc chamber
is still used to measure the absorbed dose with the appropriate CE factors.
A new method for the treatment of Mycosis Fungoides has
been developed and described. The accelerator had to be
modified to be able to deliver a beam suitable for whole
body irradiation. Previous to this method, the patient
had to be treated by thirty or more fields to cover the
whole body. Various problems were experienced where the
fields were joined. The new method uses the well known
Stanford technique of two fields angled at plus and minus
fifteen degrees to the horisontal direction. In this way
a flat field (± 2,5%) could be generated. It was also found
that a sixfield technique with the fields spaced equally
around the body, gave the best results. The various aspects of the dosimetry is described, as well as the safety measures employed to protect a patient from accidental
over exposure. This is a very real problem since the linear
accelerator has to produce a dose-rate of approximately 6500
Rads per minute at one metre from the focus to deliver a
dose-rate of 100 Rads per minute at five metres where the
whole body irradiation is done. It was found that the doserate
for electrons in air varies with distance according to
r-2,59, where r is the distance from the source.
The whole body irradiation procedure led to a study of
the physical properties of electron beams. The beams employed in radiotherapy can be classified as broad beams.
No mention of broad beam electron scattering could be found
in the literature. A detailed study was therefore made of
the scattering parameters for broad beams. The stopping
power for electrons in air was determined for the electron
energies in normal use at this institute. These results
were compared to the values predicted from the theory of
electron pencil beam scattering through thin foils. Although the results indicated that there is a reasonable
correlation between the theory and experiment, it is recommended that a more thorough theoretical study be conducted and the theory of pencil beam scattering extended
to cover broad beam scattering. The well known Monte Carlo technique could be a useful procedure to employ.
A problem which has been bothering radiotherapists, is
the reduction in depth dose when the focal to skin distance is reduced. This led to the use of the longest
practical focal to skin distance that could deliver satisfactory dose-rates. The isocentric technique, which involves an effective reduction in the f.s.d. by 15 cm, has
therefore not commonly been used on deep X-ray- and short
f.s.d. cobalt machines. This technique is however feasible
with a linear accelerator.
Accelerators are capable of delivering dose-rates of up
to five hundred rads per minute at an isocentric distance
of one metre. Furthermore, calculations and experimental evidence indicate that at an isocentric distance of
one metre there is no significant difference in the depth
dose or accuracy of the isocentric- or constant f.s.d.
By utilising the isocentric technique the treatment fields
could be altered from the control console. This would subsequently reduce setting up times and the patient would be
handled less often. Due to the reduction of patient handling the isocentric technique would be less suspectable to
error, although one error could be more serious than with
the constant f.s.d. technique. A project is envisaged where
the PDP 8 computer, which is at present being used in a DEC
Rad 8, Radiotherapy planning system, will be interfaced
with the existing linear accelerators, to act as a vigilant
to minimize human error.
It is concluded that on linear accelerators, the isocentric technique is superior to the constant f.s.d. technique.||en_ZA