Radiation Physics and Chemistry 116 (2015) 130134

Contents lists available at ScienceDirect

Radiation Physics and Chemistry

journal homepage: www.elsevier.com/locate/radphyschem

Characterization of amorphous thermoluminescence dosimeters for

patient dose measurement in X-ray diagnostic procedures
N.N.H. Ramli a, H. Salleh b, G.A. Mahdiraji c, M.I. Zulkii d, S. Hashim e, D.A. Bradley f,g,
N. M. Noor a,n
a
Department of Imaging, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia
b
Medical Physics Group, Malaysia Nuclear Agency (Nuclear Malaysia), Bangi, 43000 Kajang, Selangor, Malaysia
c
Photonics Research Group, Department of Electrical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia
d
Faculty of Engineering, Multimedia University, 63100 Cyberjaya, Selangor, Malaysia
e
Department of Physics, Universiti Teknologi Malaysia, 81310 Skudai, Malaysia
f
Department of Physics, University of Surrey, Guildford, Surrey GU2 7XH, UK
g
Department of Physics, University of Malaya, 50603 Kuala Lumpur, Malaysia

H I G H L I G H T S

We investigate the use of novel Ge-doped amorphous silica at bers as thermoluminescence dosimeters.

The investigative irradiations were made using a Toshiba X-ray machine.

Ge-doped bers represent a viable system for use in diagnostic dosimetry.

art ic l e i nf o a b s t r a c t

Article history: We investigate the use of novel Ge-doped amorphous silica at bers as thermoluminescence dosimeters
Received 15 October 2014 (TLDs) in verifying patient entrance surface-dose (ESD) in diagnostic examinations. Selected bers with
Accepted 20 January 2015 established dosimetric characteristics (including energy dependence, linearity, reproducibility, and fad-
Available online 22 January 2015
ing) were loaded into plastic capsules in groups of six. The bers have been calibrated against a parallel
Keywords: plate ionization chamber, use being made of x-rays generated at 70 kVp, accessing a Secondary Standards
Dosimetric characteristics Dosimetry Laboratory (SSDL) facility. The ber characterization measurements were made using a
Amorphous Toshiba X-ray machine operating within the nominal energies range 40 kVp to 150 kVp, for doses in the
Lithium uoride range 0.02 mGy up to 3 mGy. For doses from 2 mGy up to 150 mGy, the at bers exhibit linearity be-
Luminescence
tween TL yield and dose, reproducible to better than 3% standard deviation following repeat measure-
ments (n 3). A marked energy-dependent response is observed for photons generated at potentials
from 40 kVp to 150 kVp. From present results, it is concluded that Ge-doped bers represent a viable
system for use in diagnostic dosimetry, corrections being made for the various factors inuencing TL
yield.
& 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction The measurement of patient dose in diagnostic radiology involves

X-ray radiology provides a powerful range of clinical diagnostic
tools, with annually many millions of studies carried out world-
wide. Within the constraints of the clinical benets of the diag-
nostic procedures outweighing the risks, dosimetry has been es-
sential in evaluating the potential for harm, determining the ra-
diation dose received by the patient, and also for optimization and
justication of diagnostic examination (Meghzifene et al., 2010).
n
Corresponding author. Fax: 60 389426957.
E-mail address: noramaliza@upm.edu.my (N. M. Noor).
direct, indirect and calculational methods, with thermo-
luminescence dosimeters (TLD) being commonly utilized in direct
measurement of entrance surface dose (ESD) (Faulkner et al.,
1999). Optically stimulated luminescence (OSL) forms the basis of
yet another passive form of device that can be used in patient dose
measurement. Thus said, in accord with the International Atomic
Energy Agency (IAEA) International Code of Practice on Dosimetry
in Diagnostic Radiology, methods or devices that are to be used to
measure radiation dose need to be calibrated and to comply with
the recommended procedures for the attainment of standardized
dosimetry (IAEA, 2007).

http://dx.doi.org/10.1016/j.radphyschem.2015.01.016
0969-806X/& 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 N.N.H. Ramli et al. / Radiation Physics and Chemistry 116 (2015) 130134
In recent years there has been increasing interest in adopting
optical ber as radiation sensors in medical application, typically
being made of doped silica and responding to radiation both as a
result of the existence of intrinsic or extrinsic defects in the
medium (OKeeffe et al., 2008). Previous studies have shown
commercially available Ge-doped optical bers to have high po-
tential as thermoluminescence (TL) material for radiotherapy do-
simetry due to their low cost and small size (outer diameter of
120 mm), being useful over a wide range of incident energy, as
well as giving excellent TL response (Alawiah et al., 2013; Bradley
et al. 2011; Hashim et al., 2008; Noor et al., 2010, 2011). Present
study seeks to evaluate the dosimetric characteristics of a novel
form of optical ber fabricated to be at and irradiated using X-ray
photons at energies applied in diagnostic radiology. The response
of at optical bers are compared with standard TL and OSL do-
simeters, TLD-100 (Li:Mg,Ti) and nanoDot (Al2O3:C), respectively.
2. Materials and methods
2.1. Fabrication and designation of the at ber
The bers were fabricated using the Modied Chemical Vapour
Deposition (MCVD) method, accessing facilities at the laboratory
located at Multimedia University. The preform fabrication was
carried out using using a synthetic fused silica substrate tube
[25 mm outer diameter /19 mm inner diameter] (Fig. 1). The sub-
strate tube was etched with a ow of sulfur hexauoride (SF6) to
clean the inner surfaces followed by the deposition and sintering
of two cladding layers. The deposition of the silicagermania core
layers was then performed at 2100 C. The ow rate of oxygen (O2)
carrier to the silicon tetrachloride (SiCl4) bubbler (temperature
(T) 35 C) was kept at 200 standard cubic centimeter per minute
(sccm) while that to the germanium tetrachloride (GeCl4) bubbler
(T 35 C) was 320 sccm. Additional ow of O2 (2000 sccm) was
introduced to the MCVD main delivery line in order to enhance the
oxidation of the precursors. Helium (1000 sccm) was also fed into
the delivery line to assist in heat transfer. Throughout the whole
deposition process, the tube rotation was kept at 30 rotations per
minute (rpm) and the burner traverse speed was set to 125 mm/
min. The tube containing germanium dioxide (GeO2) in the core
was nally collapsed to a solid preform in ve stages at tem-
peratures 42200 C. During collapse, O2 was introduced to the
substrate tube to preserve its inner pressure. The O2 ow rate was
gradually reduced as the tube diameter decreased. The subsequent
pulling process was performed by members of the Photonics Re-
search Group, Department of Electrical Engineering, University of
Fig. 1. Fabrication process of optical ber.
131
Malaya, using the drawing tower located there. This has lead to the
fabrication of optical bers of various cross-sectios, with thickness
and widths of: 60  180 mm2, 85  270 mm2, 100  350 mm2,
165  620 mm2 and 200  750 mm2, all with centrally located
6 mol% Germanium dopant.
2.2. Preparation of the sample
Propanol was used to clean the ber before it was cut into
strips of 6.0 71.0 mm length using a diamond cone-point cleaver.
The mass of segmented bers were measured using an electronic
balance (Mettler Toledo, Greifensee, Switzerland). The mass was
used to normalize thermoluminescence (TL) response in terms of
TL yield per unit mass. To form a series of individual dosimeters, in
each case ten segmented bers were grouped and put into a ge-
latine capsule. To minimize surface abrasion of the bre and de-
position of dust or nger oil vacuum tweezers (Dymax 5, Surrey,
UK) were used in the handling of the optical bers. For the pur-
poses of comparison, LiF-based dosimeters were also used; these
were TLD-100 chips (3.2 mm  3.2 mm  0.89 mm). In order to
form a series of individual dosimeters, in each case three chips
were grouped and put into gelatine capsules. In regard to the OSL
dosimeters (Standard nanoDot) that were also used for compar-
ison, these were of dimension 10 mm  10 mm  2 mm. Each na-
noDot has a dot of OSL material with a diameter of 4 mm enclosed
in a light-tight carrier. In order to easily identify and track the
individual dosimeters the carriers have unique serial numbers.
Before irradiation, all dosimeters were calibrated using X-rays
generated at 80 kVp using the Secondary Standard Dosimetry Lab
(SSDL) facility located at the Nuclear Malaysia Agency.
2.3. Annealing
A furnace (Carbolite, Derbyshire, UK) was used to anneal the
bers, ensuring that any previous residual radiation history was
minimized. For this, the bers were placed in a brass holder and
annealing was performed at 400 C for 1 h. To avoid thermal
stress, the bers were retained in the brass holder for 8 h while
slowly cooling. Then, the bers were placed inside a black con-
tainer to minimize exposure to light prior to irradiation. The TLD-
100 samples were also annealed prior to irradiation at 400 C for
1 h, followed by 16 h at 80 C, then being allowed to be cooled in
the oven (Noor et al., 2012).
2.4. Sample irradiation
The sample irradiations were carried out using a model KXO-
50S (Toshiba, Shibaura, Japan) X-ray machine located in the
Medical Physics Group Laboratory at the Nuclear Malaysia Agency.
The X-ray tube of this machine has inherent beam ltration of
3 mm Al. The gelatine capsules containing the ber and TLD-100
dosimeters and the nanoDot dosimeters were placed on the sur-
face of a Perspex (Polymethyl methacrylate, PMMA) phantom at
the center of a 10  10 cm2 eld size. A focus to surface distance
(FSD) of 100 cm was employed in this study. A schematic of the
experimental set-up is shown in Fig. 2.
Samples
Field size
Solid water phantom
RaySafe detector
Fig. 2. Experimental set-up of the sample.
132 N.N.H. Ramli et al. / Radiation Physics and Chemistry 116 (2015) 130134
2.4.1. Linearity
To examine the response against dose, the dosimeters were
irradiated using a nominal tube potential of 70 kVp to doses ran-
ging from 0.02 mGy up to 3 mGy measured using a dose detector
(RaySafe, Billdal, Sweden).
2.4.2. Energy dependence
To investigate the energy dependence, irradiations were made
using identical doses of 3 mGy for a range of nominal tube po-
tentials, from 40 kVp until 150 kVp, using the previously men-
tioned Toshiba X-ray machine and the 1.25 MeV mean energy of
gamma rays from a 60Co source located at the SSDL facility. Due to
the polyenergetic nature of diagnostic X-ray beams, effective en-
ergy (keV) is typically used to describe the mean energy. Typically
diagnostic X-ray tube beams are estimated to have an effective
energy of the order of one-third to one-half the maximal values
(Bushberg et al., 2011). As such, using the one-half characterisa-
tions, the effective energies provided at accelerating potentials of
40 kVp, 70 kVp, 100 kVp and 150 kVp would be 20 keV, 35 keV,
50 keV and 75 keV respectively.
2.4.3. Reproducibility
To investigate reproducibility the dosimeters were irradiated
with a tube potential of 60 kVp, 80 kVp, 100 kVp and 150 kVp and
a dose of 1 mGy. The irradiation was repeated using the same -
bers, subsequent to re-annealing on each occasion. The standard
deviation percentage of ve ber readings for each irradiation
have been calculated.
Fig. 3. Linearity of at bers and commercial ber with dose (A) within 2 mGy to 150 mGy, (B) within 2 mGy to 10 mGy [this is enlarged from (A)].
2.4.4. Fading
Signal loss over a period of time was observed by irradiating
the dosimeters at a nominal tube potential of 80 kVp and dose of
1 mGy. Post-irradiation, and prior to readout, the dosimeters have
been kept in a black box and stored at room temperature.
2.5. Readout
The at ber samples and TLD-100 chips were read out using a
Harshaw 3500 (Thermo Fisher Scientic Inc, USA) TLD reader, with
a nitrogen gas atmosphere of 0.5 bars. The timetemperature
prole (TTP) used during readout was: preheat temperature of
180 C for 5 s; readout temperature of 400 C for 13 s and heating
cycle rate of 25 C s  1. An internal annealing temperature of
400 C was used for 5 s to remove any residual signal in the bers.
The nanoDot dosimeters were read out using Microstar Reader
(Landauer Inc, Glenwood, IL). A low-intensity light emitting diode
(LED) beam was used for irradiation measurements, the reader
operating in continuous-wave light modes with a 1 s illumination-
read period (Jursinic, 2007).
3. Results and discussion
3.1. Linearity
Fig. 3 shows the relationship between TL yield and dose for a
commercial single-mode cylindrical cross-section ber with a
central 9 mm doped core and also for the various at bers. The TL
 N.N.H. Ramli et al. / Radiation Physics and Chemistry 116 (2015) 130134 133

yield with dose is highly linear for the at bers (r2 4 0.99), in
particular from 2 mGy up to 150 mGy, being reproducible to better
than 3% standard deviation following repeat measurements (n 3).
Also apparent is that the TL yield is superior for the at bers and
in particular the sensitivity to radiation dose increases with in-
crease in the cross-section of the bers.
Throughout the dose range 2 mGy to 150 mGy, the largest at
ber (200  750 mm2) gives the greatest TL response, followed by
that for the 165  620 mm2, 100  350 mm2, 85  270 mm2, and
60  180 mm2 at bers. In assessing dose linearity, the at ber of
cross-sectional size 165  620 mm2 was chosen as the best ex-
ample, with a superior r2 value of 0.9993. The dose response of the
selected at bers (FF), TLD-100 and nanoDot were evaluated by
Fig. 5. The energy dependence of FF and TLD-100 to the range of nominal tube
plotting the graph of Linearity Index function, f (D) against dose, as potential and mean energy of Co-60.
shown in Fig. 6. The equation is that rst proposed in a study of TL
nonlinearity in dose dependence (Chen and McKeever, 1994), as
follows:
S (D) So S (D1) So
f (D ) = /
D D1

where

S0 intercept of the TL intensity axis,

S(D) TL response corresponding to dose D,
D1 dose in the linear region,
S(D1) TL response corresponding to dose D1,
with f(D)1 indicating linearity, f(D)41 supralinearity, and
f(D) o1 sublinearity.

Based on Fig. 4, the doseresponse of the bers are sublinear at

doses lower than 0.6 mGy and predominantly supralinear for do- Fig. 6. Fading of FF, TLD-100 and nanoDot within 14 days.
ses from 0.6 mGy up to 2 mGy; beyond this the dose response is
linear. For TLD-100, a supralinear doseresponse is indicated at
0.02 mGy and sublinear for doses of 0.04 mGy up to 0.6 mGy. The
Table 1
nanoDot also shows a predominantly sublinear doseresponse at
Dose reproducibility at 60, 80, 100 and 150 kVp.
doses lower than 0.6 mGy. For all of the dosimeters compared
herein, it can be seen from Fig. 4 that the doseresponses is linear Nominal tube poten- Dose No. of TL yield per unit mass (nC/g)
for doses 41 mGy. tial (kVp) (mGy) ber
Mean SD SD (%)

60 0.955 5 2123.66 58.19 2.74

3.2. Energy dependence 80 0.990 5 2285.00 16.23 0.71
100 1.064 5 2131.33 62.50 2.93
Sensitivity as a function of energy has been investigated for the 150 0.957 5 2101.46 30.08 1.43
at bers and TLD-100 (Fig. 5). Both forms of dosimeter show a
similarly characteristic trend of decreasing sensitivity with photon
3.3. Reproducibility
energy, as expected for an increasingly inefcient detection pro-
cess. The more signicant decrease seen for the at bers is also as Table 1 demonstrates the mean TL yield per unit mass for
expected, note being taken of the greater effective atomic number corresponding doses and number of segmented ber readouts. At
of the silica bers compared to TLD-100. the nominal delivered dose of 1 mGy, the standard deviation of the
bers was observed to be between 0.7 and 2.9 %.

3.4. Fading

Fig. 6 demonstrates the signal loss of the at bers, TLD-100

and nanoDot as a function of time post-irradiation. All dosimeter
responses were normalized to the response at day three post-ir-
radiation. Over 14 days, the rate of fading was 24 % for the at
bers, 26 % for TLD-100 and 2 % for nanoDot.

4. Conclusion

The at ber shows a linear doseresponse for doses 41 mGy.

Fig. 4. Plot of Linearity Index f (D) against dose of FF, TLD-100 and nanoDot. For these bers there is also a marked change in energy-
134 N.N.H. Ramli et al. / Radiation Physics and Chemistry 116 (2015) 130134
dependence for X-rays delivered at tube potentials between
40 kVp and 150 kVp. Reproducibility of better than 3.0% was also
found. In conclusion, with correction factors applied for the var-
ious dosimetric characteristics investigated herein (energy de-
pendence, linearity, reproducibility and fading), the at bers
show themselves to be a feasible form of dosimeter for use in
diagnostic radiology.
Acknowledgment
The authors would like to acknowledge PUTRA Grant no:
9404200 from Universiti Putra Malaysia, Universiti Malaya High
Impact Research Grant: UM.C/625/3/HIR/MOHE/SC/33, A 000007-
50001 and Malaysia Fundamental Research Grant Scheme (FRGS)
No: 02-02-13-1317FR from Ministry of Education, Malaysia for
supporting this research. We are thankful for the use of irradiation
facilities at the Malaysian Nuclear Agency.
References
Alawiah, A., Intan, A.M., Bauk, S., Abdul Rashid H.A., Yusoff Z., Mokhtar, M.R., Wan
Abdullah, W.S., Mat Sharif, K.A., Mahdiraji, G.A., Mahamd Adikan, F.R., Tamchek,
N., Noor N.M., Bradley D.A., 2013. Thermoluminescence characteristics of at
optical ber in radiation dosimetry under different electron irradiation condi-
tions. In: SPIE Proceedings of the Micro-structured and Specialty Optical Fibres
II, 87750S, vol. 8775 (May 3, 2013).
Bradley, D.A., Hugtenburg, R.P., Nisbet, A., Abdul Rahman, A.T., Issa, F., Noor, N.M.,
Alalawi, A., 2011. Review of doped silica glass optical bre: their TL properties
and potential application in radiation therapy dosimetry. Appl. Radiat. Isot. 71,
211.
Bushberg, J.T., Seibert, J.A., Leidholdt Jr., E.M., Boone, J.M., 2011. The Essential Phy-
sics of Medical Imaging, 3rd ed North American: Lippincott Williams & Wilkins,
USA.
Chen, R., McKeever, S.W.S., 1994. Characterization of nonlinearities in the dose
dependence of thermoluminescence. Radiat. Meas. 23, 667673.
Faulkner, K., Broadhead, D.A., Harrison, R.M., 1999. Patient dosimetry measurement
methods. Appl. Radiat. Isot. 50, 113123.
Hashim, S., Ramli, A.T., Bradley, D.A., Wagiran, H., 2008. The thermoluminescence
response of Ge-doped optical bres to x-ray photon irradiation. J. Phys. UTM 3,
3137.
IAEA, 2007. Dosimetry in Diagnostic Radiology: An International Code of Practice.
International Atomic Energy Agency (IAEA), Vienna, Austria (Technical Report
Series No. 457).
Jursinic, P.A., 2007. Characterization of optically stimulated luminescent dosi-
meters, OSLDs, for clinical dosimetric measurements. Med. Phys. 34 (12),
45944604.
Meghzifene, A., Dance, D.R., Mclean, D., Kramer, H., 2010. Dosimetry in diagnostic
radiology. Eur. J. Radiol. 76 (1), 1114.
Noor, N.M., Hussein, M., Bradley, D.A., Nisbet, A., 2010. The potential of Ge-doped
optical bre TL dosimetry for 3D verication of high energy IMRT photon
beams. Nucl. Instrum. Methods Phys. Res. A 619, 157162.
Noor, N.M., Hussein, M., Bradley, D.A., Nisbet, A., 2011. Investigation of the use of
Ge-doped optical bre for in vitro IMRT prostate dosimetry. Nucl. Instrum.
Methods Phys. Res. A 652 (1), 819823.
Noor, N.M., Shukor, N.A., Hussein, M., Nisbet, A., Bradley, D.A., 2012. Comparison of
the TL fading characteristics of Ge-doped optical bres and LiF dosimeters.
Appl. Radiat. Isot. 70, 13841387.
OKeeffe, S., Fitzpatrick, C., Lewis,, E., Al-Shammaa, A.I., 2008. A review of optical
bre radiation dosimeters. Sens. Rev. 28 (2), 136142.