Commissioning of an Optically-Stimulated Luminescence

(OSL) system for remote dosimetry audits

J.F. Aguirre, P. Alvarez, C. Amador, A. Tailor, D. Followill, G. Ibbott
Department of Radiation Physics
The University of Texas, M.D. Anderson Cancer Center, Houston, Texas 77030

Purpose Dose Determination Algorithm Results (cont’d) Results (cont’d)

For 30 years the Radiological Physics Center The response of an OSL dosimeter is quasi linear with dose ELEMENT CORRECTION FACTORS (ECF)(cont’d) ENERGY/BLOCK CORRECTION
(RPC) has used TLD dosimeters for remote for any energy. Small deviations from linearity are corrected
audits of beam output of photon and electron Table 1: Average ECF after nine cycles Preliminary results showed that the signal per unit dose was
together with loss of signal from the difference in time
beams and energy checks for electron beams. largely independent of energy when irradiations were
between irradiation and reading as well as corrections for Dosimeter ID AVG ECF STDEV
Acrylic blocks with capsules containing TLD performed in a full phantom.5 Dosimeters in acrylic blocks
energy/block and position of the OSL in the beam.
powder are sent for each beam. The powder is DN09305639P 1.035 0.34% were irradiated with different beam energies to a measured
used as a disposable dosimeter. The RPC has DN09307843U 0.950 0.50% dose of 100 cGy and read in one session together with
previously described the use of the remote
Methodology standards irradiated in the cobalt beam at the same dose level.
DN09307865O 0.989 0.83%
audits to identify units with beam Some energy dependence combined with a correction for
DN09307916P 0.974 0.85% reduced backscatter was determined as:
measurements exceeding 5% and 5 mm. The
X-ray or e beam DN09308972Q 1.045 1.34%
(cGy/Signal )energyE
system uncertainty is 1.5% (1 standard cobalt
deviation) indicating high confidence in the DN093090941 0.997 0.24%
Signal: KE =
5% threshold for acceptability.1,2
Reading
DN09309159T
DN09309249S
1.010
1.030
0.60%
0.48%
(cGy/Signal )Cobalt
Optically-Stimulated Luminescence (OSL) corrected for
dosimetry with aluminum oxide doped with fading, linearity DN09309355X 1.012 0.39% The signal was corrected for fading, linearity and the sensitivity
carbon has been extensively used to monitor and energy DN09309697J 0.989 0.96% of each dosimeter (ECF). KE for any energy E relative to a cobalt
personal occupational radiation dose and the beam is shown below:
use of OSL dosimeters for dose SIGNAL REPRODUCIBILITY
measurements at therapeutic levels has been Sensitivity (S) = Dose / signal Dose = S x signal Energy KE
studied in the last few years.3,4 The RPC When dosimeters were irradiated and read as described 6 MV 0.992
performed a promising initial evaluation of a Formulism earlier, the corrected readings demonstrated a standard
15 MV 1.028
commercial system using the microStar deviation of 0.8%.
18 MV 1.044
System™ with InLight™ dosimeters5 and Dose = S · signal ·ECF · DCF • KL · KF · KE NON LINEARITY
decided to purchase and commission 5e 1.036
dosimeters and instrumentation with the goal Dosimeters irradiated at doses between 25 cGy and 350 6e 1.027
ECF: Element correction factor cGy showed supralinear responses. A correction was
of implementing an OSL-based system into its 7e 1.026
DCF: Depletion correction factor applied to compensate for this supralinearity. That
remote audit program. The system was 8e 1.033
KL Supra linearity correction correction was:
available with the InLight nanoDot™
KF: Fading correction 9e 1.019
dosimeter.
KE: Energy/block correction 10e 1.023
Materials 12e 1.017
OSL reader microStar System™ (2 units) Results (cGy/Signal)DoseD
15e 1.025
InLight nanoDot™ dosimeters KL = 16e 1.014
Annealing light box from Landauer
SIGNAL CAPTURE
(cGy/Signal)100cGy 20e 1.019
Cobalt beam A reading time of 7 seconds was adopted from studies by
6 – 18 MV photon beams Homnick5. The reference dose was 100cGy. 1.06 Conclusions
5 – 20 MeV electron beams
Repetitive readings of a single dosimeter with two different

KL Normalized to 100 cGy

Photon mini-phantom blocks for in-air 1.04 1. The measurements reported here demonstrate that OSL
readers showed predictable behavior that was reader- dosimetry is an acceptable alternative for remote dosimetry of
irradiations of dosimeters y =-0.028618Ln( x) + 1.132293
dependent and represented a loss of around 0.2% per 1.02
teletherapy beams.
Electron phantom blocks for full phantom
reading. This study led to the conclusion that more than y =- 0.0286036Ln( x) + 1.1317246
irradiations of dosimeters 1.00
2. The reproducibility reported here was achieved by irradiating
one reading was needed and that three readings provided
acceptable confidence. The spread over three readings was 0.98
two dosimeters at each point and taking three readings from
such that no depletion correction was needed. The graphs each dosimeter. The standard deviation of the three readings
below show the differences in quality of the signal between 0.96 should be better than 1.5%; if not, more readings can be taken
two readers. 0.94
and depletion corrections applied.
0 50 100 150 200 250 300 350 400
3. The response of the dosimeters used for this study deviated
Dose (cGy) from the average by up to 8%. A relative dose response
correction was found necessary for each dosimeter.
1.2
Depletion
Methods 3
Polynomial
2
y = -4E-08x + 1E-05x - 0.0025x + 1.0053
2
R = 0.9913
Exponential
y = 0.995e
2
-0.0018x

R = 0.9888
The KL function was determined from multiple
4. It was observed that the characteristics of individual dosimeters
were independent of the number of irradiation/anneal cycles as
The following properties weredetermined as 1
long as the cumulative dose was less than 1000 cGy.
irradiations of a single dosimeter and was reproducible
part of the characterization of the system:
as long as the cumulative dose to the dosimeters did not 5. The need to apply ECF corrections and the limit on cumulative
0.8 exceed 1000 cGy. Consequently, 1000 cGy was identified
Normalized reading

READER dose for each dosimeter required tracking the dose received by
as the limit of cumulative dose. each dosimeter.
• Stability
• Reading cycle 0.6
FADING 6. Uniformity of the dosimeter response and depletion rate of the
DOSIMETER 3
Polynomial
2
y = -4E-08x + 2E-05x - 0.0044x + 1.0017
Exponential
y = 0.9828e
-0.0035x
signal can be affected by the characteristics of each reader. The
• Depletion rate 2
R = 0.9839 2
R = 0.9816 Dosimeters were irradiated at different dates to a dose of
0.4
methodology used here achieved a distribution in the
• Dependence of depletion rate with 100 cGy and read in one session. Up to 4% of the signal
measurements of less than 1% (SD) after the necessary
reader was was lost over a period of 120 days with the decrease
0.2 corrections. This may be a criterion for acceptance of a reader.
• Cumulative dose limit being more pronounced in the first few days. In practice,
Reader 1 Reader 2
• Number of readings per dosimeter dosimeters will be read approximately 7 days after
• Relative dose response or element 0
0 20 40 60 80 100 120 140
irradiation. The graph below shows KF values normalized References
correction factor (ECF) Number of readings
at two days. The tests reported here used a delay of 2 or [1] Aguirre JF, Tailor R, Ibbott G, Stovall M and Hanson, W. Thermoluminescence
• Variability of ECF with reader more days after irradiation. dosimetry as a tool for the remote verification of output for radiotherapy beams: 25 years
• Variability of ECF with dose of experience. In: Proceedings of the International Symposium on Standards and Codes
ELEMENT CORRECTION FACTORS (ECF) FADING CORRECTION Normalized 2 days of Practice in Medical Radiation Dosimetry IAEA-CN-96/82, Vienna: IAEA. 2002; 191–
• Dose linearity correction
A set of dosimeters went through a series of 100 cGy 99.
• Signal fading correction 1.015
[2] Kirby TH, Hanson WF and Johnston DA. Uncertainty analysis of absorbed dose
• Energy/block correction irradiations, reading and annealing cycles and the relative 1.010
1.005 -0.0072 calculations from thermoluminescence dosimeters. Med Phys 1992; 19:1427–1433.
ANNEALING signal factors were calculated and compared with the y = 1.005x
Fa ding C orr e c tion

1.000
2 [3] R. C. Yoder and M. Salasky. Optically stimulated luminescense dosimeters – an
• Optimal annealing time and average signal. The factor for each dosimeter was constant 0.995 R =1
alternative to radiological monitoring films. Proceedings of the Optical Engineering
0.990
recommended instrumentation. within a standard deviation of 0.3 to 1.3%. See Table 1. Midwest 95, Illinois Institute of Technology, 18 May 1995.
0.985
• Variability of ECF with annealing The exercise was also performed with doses of 25 and 300 0.980 [4] P.A. Jursinic. “Characterization of optically stimulated luminescent dosimeters,
0.975 OSLDs, for clinical dosimetry measurements”. Med Phys 2007; 34:4594–4604.
cGy. In addition, readings were performed on the second 0.970
[5] J. Homnick. G. Ibbott, A. Springer, J. Aguirre. “TH-D-352-05:Optically Stimulated
reader without any significant change. 0.965
Luminescence (OSL) Dosimeters Can Be Used for Remote Dosimetry Services”. Med
0.0 20.0 40.0 60.0 80.0 100.0 120.0 140.0
Days Elapsed Phys 2008; 35:2994–2995.
The investigation was supported by PHS grants CA10953 and CA81647
awarded by the NCI, DHHS.