Absorbed dose to water dosimetry for heavy charged particle beams
using solid-state luminescence dosimeters
固体蛍光線量計を用いた重荷電粒子線 水吸収線量計測に関する研究
2016 年 09 月
首都大学東京大学院
人間健康科学研究科 博士後期課程 人間健康科学専攻
放射線科学域 学修番号：13997607
氏 名：張 維珊
（指導教員名： 齋藤 秀敏 ）
博 士 学 位 論 文
（要約）
Abstract
As heavy charged particle (HCP) therapy facilities increased, the demand of dosimetry system
for HCP beams increased. Radiophotoluminescence glass dosimeter (RGD) is widely used in
photon radiotherapy because it offers the advantage of good reproducibility, dose linearity, and
unlimited repeatable readings. Besides point dose measurement, two-dimensional dose
measurement is also desirable for dose verification in radiotherapy. Recently, a slab-type
thermoluminescence dosimeter (TLD) based on the thermoluminescence (TL) phosphor Li3B7O12:Cu
has been developed. Some reports have concerned with the feasibility of the above-mentioned
solid-state luminescence dosimeters (SSLDs) in HCP beams. However, the problem of LET
dependent response have not been solved. To use the SSLDs in HCP dosimetry, comprehensive
investigation and a method taking the LET dependence into account are necessary.
To use RGDs in proton dosimetry, a method considering the LET dependent response and the
change of beam quality was proposed. The LET dependence of radiophotoluminescence (RPL)
efficiency ekRGD,γ have been investigated experimentally and the change of beam quality have been
calculated by Monte Carlo simulation. For practical use, the residual range Rres was used as a
quality index to determine the correction factors. The proposed method for RGD in proton beam
was evaluated by comparing absorbed dose to water Dw measured by RGD DwRGD and ionization
chamberDwIC at difference depth in therapeutic proton beams. As a result, for both non-modulated
and modulated beam, DwRGDshowed a good agreement with DwIC except where Rres is less than 1 cm.
In conclusion, following the proposed method, RGD can be used as dose verification tool and postal
audit dosimeter for proton therapy.
In regards to RGDs in carbon dosimetry, the LET dependence of ekRGD,g has been investigated
using carbon beam and the change of the fraction of secondary particles between non-modulated and
modulated carbon beam has been calculated using Monte Carlo simulation. It is clarified that ekRGD,g
estimated by proton beam is lower than that by carbon beam even they have same LET and the
change of the fraction of secondary particle needs to be consider when using RGD in carbon
dosimetry.
A high temperature ratio (HTR) method has been proposed to estimate LET and correct TL
efficiency. To realize using Li3B7O12:Cu in carbon dosimetry, the HTR method has been
considered, and slow heating method was used to improve the reproducibility of the glow curve.
The relation between TL efficiency ekTLD,γ , relative HTR HTRk,g, and LET for Li3B7O12:Cu were
investigated and the ekTLD,γ as a function of HTRk,g was derived. The feasibility of the HTR
method for Li3B7O12:Cu was evaluated by comparing the Dw by Li3B7O12:Cu DwTLD andDwIC in the
modulated carbon beam. After the HTR correction, the maximum of dose difference between
IC w TLD
w andD
D was decreased from 50 % to 30 %. It is concluded that the accuracy of Dw
estimation in carbon beams was improved by HTR method, and the usefulness of slow heating
method was revealed which is helpful for Li3B7O12:Cu in HCP dosimetry.
In this thesis, with an aim of using SSLDs in HCP dosimetry, the methods that take into
account the LET dependence were proposed and the feasibility were evaluated.
Overview of the thesis
Accurate dosimetry is a fundamental requirement for radiotherapy. There have been
several reports concerned with the linear energy transfer (LET) dependence of passive type
solid-state dosimeters and the application in heavy charged particle (HCP) beams. However, none
of them contains a comprehensive discussion for measuring absorbed dose to water Dw in HCP
radiotherapy.
Radiophotoluminescence glass dosimeter (RGD) was chosen in this thesis because the
following reasons: 1) the small volume of RGD is suitable for being an in-vivo dosimeter, 2) the
characteristics of less fading effect, small variation in individual sensitivity and repeatable readout
are suitable for postal dosimetry audit of radiotherapy, 3) the radiophotoluminescence (RPL)
efficiency decreases continuously with increasing LET which means it is easy to correct compare to
other commercial available solid-state luminescence dosimeter, 4) it is used as a postal dosimeter for
photon radiotherapy in Japan.
On the other hand, since the dose distribution become more complex in modern radiotherapy,
two-dimensional (2D) dose measurement is recommended for ensuring that the treatment machines
are delivering the expected dose distributions to each patient. Recently, a slab-type
thermoluminescence dosimeter (TLD) based on the thermoluminescence (TL) phosphor Li3B7O12:Cu
has been developed. The TL phosphor Li3B7O12:Cu showed good dose linearity and excellent
spatial resolution. However, it has been reported that the TL efficiency of the phosphor
Li3B7O12:Cu is LET dependent.
To estimate Dw using above-mentioned dosimeters in HCP beams, this thesis will clarify the LET quenching effect of detector and propose protocols for each dosimeter with an attempt to solve
the LET dependent response. This thesis is composed of seven chapters and the content of each
chapter is as follows:
Chapter 1 Introduction:
The role of radiotherapy in cancer treatment, and current status of HCP therapy were briefly
described. And the literatures concerned with the problems of dosimeters in HCP beams was
reviewed. Finally the purpose of this thesis was described.
Chapter 2 Background knowledge:
An essential pre-requisite knowledge for this study has been introduced in this chapter.
Firstly, the radiation quantities related in this thesis and the interactions between HCPs and matter
was outlined. Secondly, the theories related dosimetry and detector response in HCP beams and the
major dosimetry protocol for Dw were summarized. Finally, the principles of solid-state
luminescence dosimeters (RGDs and TLDs) were introduced.
Chapter 3 Experimental facilities:
The experimental facilities including the facilities of irradiation and the dosimetry systems
used in this thesis are introduced in this chapter.
The weight composition of RGD (GD-302M, Asahi Techno Co., Japan) is as follows: P
(31.55 %), O (51.16 %), Al (6.12 %), Na (11.00 %), and Ag (0.17 %). The size of RGD is 1.5 mm
in diameter and 12 mm in length and the density is 2.61 g cm-3.
The TLD used in this thesis is an in-house developed TLD: TL phosphor Li3B7O12:Cu. It is
produced by combining Li2B4O7 and B2O3 at 3:1 ratio with the addition of 0.1 % CuO by weight.
The density and the effective atomic number for Li3B7O12:Cu was 1.01 g cm-3 and 7.42, respectively.
In this thesis, g-ray from a 60Co unit at National Institute of Radiological Sciences (NIRS)
was used to determine
Q0
w,
ND, for RGD and TLD and to improve the accuracy, i.e. reduce the
uncertainty,
Q0
w,
ND, was determined individually for each RGD and TLD.
Chapter 4 Estimation of absorbed dose to water in therapeutic proton beam using RGD:
A feasibility study of using RGD to estimate absorbed dose to water Dw in therapeutic proton beam. In this chapter, to realize using RGD in proton dosimetry, a method considering the change
of beam quality and the LET quenching effect in RPL efficiency ek,RGDγ was proposed.
The relationship between LET and ek,RGDγ has been experimentally investigated and the
change of beam quality, stopping power ratio of water to RGD (S r)w,RGD, has been clarified using
Monte Carlo simulation toolkit Geant4. For practical use, the residual range Rres was proposed to
be the parameter for obtaining the correction factor for ek,RGDγ and (S r)w,RGD. The calculation
of (S r)w,RGD was carried out for non-modulated and modulated proton beam to clarify if it is
suitable for RGD to use Rres as the beam quality index. Finally, the feasibility of RGD in proton
dosimetry was verified by comparing Dw by RGD DwRGD and that by ionization chamberDwIC at
difference depth in therapeutic proton beams (non-modulated and modulated) at Nagoya proton
therapy center.
Difference of (S r)w,RGD between non-modulated and modulated proton beam is small for
each Rres. Therefore, it is able to express (S r)w,RGD with single function of Rres for
non-modulated and modulated proton beams. According to the results of depth dose distribution
measurement, for both non-modulated and modulated proton beam, DwRGD showed good agreement
with DwIC except at the region of Rres < 1 cm. Therefore, it is concluded that RGD can be applied
in proton dosimetry using the method proposed in this thesis.
Chapter 5 Investigation of the feasibility of RGD in carbon beam:
Since the nuclear reaction is more significant in carbon beam, it is necessary to investigate
the degree of the particle dependence of ek,RGDγ before using RGD in carbon dosimetry. As a
pre-study for using RGD to estimate Dw in carbon beam, it is necessary to investigate the degree of the particle dependence of ek,RGDγ before using RGD in carbon dosimetry. The LET dependence of
RGD γ
ek, in carbon beam, the degree of particle dependence of ek,RGDγ and the fraction of particle with
atomic number from 1 to 6 i.e. secondary particle as a function of depth for non-modulated and
modulated carbon beam have been investigated.
The LET dependence of ek,RGDγ was experimentally investigated and the degree of particle
dependence of ek,RGDγ was evaluated evaluated by comparing ek,RGDγ obtained using carbon beam to
proton beam. The fraction of the particles with low atomic number at difference depth for
non-modulated and modulated carbon beam was calculated using Monte Carlo simulation.
As a result, ekRGD,g estimated by proton beam is lower than that by carbon beam while they
have same LET. And the particle fraction of atomic number from 1 to 6 in the non-modulated
carbon beam is different from that in the modulated carbon beam. It is concluded that the
consideration of the LET dependence and particle dependence in ekRGD,g and more investigation on
the energy distribution of the secondary particles are necessary for RGD in carbon dosimetry.
Chapter 6 Estimation of absorbed dose to water in therapeutic proton beam using a Li3B7O12:Cu based TLD:
A feasibility study of using TLD based on TL phosphor Li3B7O12:Cu to estimate Dw in
therapeutic carbon beam. To correct the LET dependence of TLD, a method using the LET
dependence in glow curve and TL efficiency ek,TLDγ named high temperature ratio (HTR) method has
been introduced. The relationship between the LET, ek,TLDγ and glow curve of Li3B7O12:Cu have
been investigated and the feasibility of HTR method for Li3B7O12:Cu in therapeutic carbon beam was
evaluated. Besides, the usefulness of slow heating method in improving the reproducibility, i.e. the
accuracy of HTR method, was also evaluated.
As a result, the usefulness of slow heating method was revealed by decreasing the coefficient
of variation of the Li3B7O12:Cu’s HTR. After HTR correction, the max dose difference between Dw
by TLD DwTLD and DwIC decrease from 50 % to 28 %. To clarify the reason for the dose
difference between DwTLD and DwIC, LET in the modulated carbon beam was calculated by Geant4
and the calculated LET was used to derive the ekTLD,g at each depth. The max dose difference
between DwTLD and DwIC decrease to 10 % using the ekTLD,g by calculated LET. Therefore, it is
considered that the dose difference resulted from the uncertainty of the relationship between HTR
and LET and it may be settled with a TL phosphor with high HTR sensitivity.
Chapter 7 Conclusion:
All the results and finding of the work in this thesis and the possible future works were
summarized in this chapter. In regards to RGD, it is concluded that following the proposed method
RGD is useable in proton dosimetry, but it is not suitable in carbon dosimetry because its own strong
particle dependence. In regards to Li3B7O12:Cu based TLD, based on our results, slow heating rate
method can improve the reproducibility of glow curve and the accuracy of HTR method. In
conclusion, HTR method improved the accuracy of Li3B7O12:Cu based TLD, but more improvement
on the sensitivity of HTR is necessary.
