«Characterization of MOSkin detector for in vivo skin dose measurement during megavoltage radiotherapy Wei Loong Jong,1 Jeannie Hsiu Ding Wong,2 Ngie ...»
JOurnAL OF APPLIEd CLInICAL MEdICAL PHYSICS, VOLuME 15, nuMBEr 5, 2014
Characterization of MOSkin detector for in vivo skin dose
measurement during megavoltage radiotherapy
Wei Loong Jong,1 Jeannie Hsiu Ding Wong,2 Ngie Min Ung,1a Kwan
Hoong Ng,2 Gwo Fuang Ho,1 Dean L. Cutajar,3 Anatoly B. Rosenfeld3
Clinical Oncology Unit,1 Faculty of Medicine, University of Malaya, Kuala Lumpur,
Malaysia; Department of Biomedical Imaging and University of Malaya Research
Imaging Centre (UMRIC),2 Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia; Centre for Medical Radiation Physics (CMRP),3 University of Wollongong, NSW, Australia email@example.com Received 28 December, 2013; accepted 15 May, 2014 In vivo dosimetry is important during radiotherapy to ensure the accuracy of the dose delivered to the treatment volume. A dosimeter should be characterized based on its application before it is used for in vivo dosimetry. In this study, we characterize a new MOSFET-based detector, the MOSkin detector, on surface for in vivo skin dosimetry. The advantages of the MOSkin detector are its water equivalent depth of measurement of 0.07 mm, small physical size with submicron dosimetric volume, and the ability to provide real-time readout. A MOSkin detector was calibrated and the reproducibility, linearity, and response over a large dose range to different threshold voltages were determined. Surface dose on solid water phantom was measured using MOSkin detector and compared with Markus ionization chamber and GAFCHROMIC EBT2 film measurements.
Dependence in the response of the MOSkin detector on the surface of solid water phantom was also tested for different (i) source to surface distances (SSDs);
(ii) field sizes; (iii) surface dose; (iv) radiation incident angles; and (v) wedges.
The MOSkin detector showed excellent reproducibility and linearity for dose range of 50 cGy to 300 cGy. The MOSkin detector showed reliable response to different SSDs, field sizes, surface, radiation incident angles, and wedges. The MOSkin detector is suitable for in vivo skin dosimetry.
PACS number: 87.55.Qr Key words: MOSFET, in vivo dosimetry, surface dose, skin dose, characterization, quality assurance (QA) I. IntrOduCtIOn Quality assurance (QA) in radiotherapy is very important in order to ensure the correct functioning of all components in radiotherapy, from treatment planning to the delivery of the treatment.(1) Nowadays, advanced radiotherapy techniques, such as intensity-modulated radiotherapy (IMRT), require patient-specific QA to be performed to ensure the accuracy of radiation delivery during radiotherapy. However, these QA and verification proceduresmay not be sufficient to ensure the accuracy of the entire radiotherapy treatment.
A number of incidents have been reported recently.(2-4) Human errors and systematic errors contributed to these incidents. Therefore, towards that end, in vivo dosimetry can detect major errors during the delivery of radiotherapy. It also can access clinical relevant differences between planned and delivered dose, record the dose received by the patient, and fulfill legal requirements.(1) a Corresponding author: Ngie Min Ung, Clinical Oncology Unit, Faculty of Medicine, University of Malaya, 50603 Kuala Lumpur, Malaysia; phone: (+60)3 7949 2456; fax: (+60)3 7960 3072; email: firstname.lastname@example.org 121 Jong et al.: Characterization of MOSkin 121 In radiotherapy, in vivo dosimetry means the measurement of the radiation dose received by a patient during treatment.(1) Ideally, a dosimeter should be positioned at the point of interest inside a patient’s body. However, in many cases it is not possible to place a dosimeter inside a real patient’s body. Hence, the placement of a dosimeter on the surface of the patient’s body becomes an alternative.
An ideal in vivo dosimeter should possess the following characteristics: (i) tissue equivalent;
(ii) small in physical size and has small sensitive volume; (iii) features (e.g., temperature, energy) which are consistent and characterizable; (iv) does not perturb the radiation field; (v) nonhazardous to humans; and (vi) able to provide real-time dosimetric information. Thermoluminescence dosimeter (TLD)(5-7) is small in size, but requires a long series of pre- and postirradiation process. Radiochromic film(8-10) has excellent dosimetric spatial resolution, is able to provide two-dimensional (2D) dosimetric information(11) and is easy to use, but it is not done in real-time and may be affected by improper handling and scanner performance. Semiconductor detectors such as diode(12-13) and metal oxide semiconductor field effect transistor (MOSFET)(14-19) are able to achieve excellent spatial resolution with their small sensitive volumes. However, the energy, angle, temperature, and dose-rate dependence of semiconductor detectors require rigorous characterization.
The dose deposited on a phantom or patient surface mainly comes from primary photon beam, backscattered radiation from the phantom, as well as radiation contamination from the accelerator. Radiation contamination arises from: (i) treatment head materials and (ii) treatment setup parameters such as source-to-surface distance (SSD), field size, and beam modifier to the surface dose.(20) These contaminations will affect the dose in the buildup region. Therefore, it is essential to determine and know the effect of these treatment parameters.
Different terminologies, such as surface dose, skin dose, and entrance dose, have been used to describe the dose measured on the surface of a phantom or a human. The definitions for these terminologies differ according to the point of measurement on the patient or phantom. Surface dose is defined as the dose on the surface of the phantom or human, which is the interface between the air and the surface. Skin dose is defined as the dose at the depth of 0.07 mm.(21) Entrance dose is defined as the dose given by the entrance beam at the depth of maximum dose.(22) Characterization of a dosimeter is normally performed at a condition where charged particle equilibrium (CPE) condition exists.(12-13,19) However, for in vivo skin dosimetry, the dosimeter should be characterized on the surface instead of the depth of maximum dose. This is because the dosimetric condition of skin surface and buildup region is different from the dosimetric condition at the depth of maximum dose. At the interface of two media (air and human tissue), CPE does not exist and there is a steep dose gradient in the buildup region. Therefore, characterization of a dosimeter on surface is needed prior to using it for in vivo skin dosimetry.
A MOSFET-based dosimeter, the MOSkin detector was designed and prototyped by the Center for Medical Radiation Physics (CMRP) in the University of Wollongong (UoW). The advantages of the MOSkin detector, such as being small in size with submicron dosimetric volume which provides excellent dosimetry spatial resolution, as well as the ability to provide real-time reading and instant readout, make it suitable for in vivo skin dosimetry measurement.
The MOSkin detector has been characterized and been used for dose measurement in megavoltage radiotherapy and brachytherapy.(23-30) In this paper, a full characterization of the MOSkin detector on the surface of a phantom simulating the actual condition for in vivo skin dosimetry (where non-CPE condition exist) was performed and reported. These include: (i) detector calibration, linearity, reproducibility;
(ii) source to surface distance dependence; (iii) field size dependence; (iv) surface dose measurement; (v) angular dependence; and (vi) wedge response. Comparison and verifications were made with previous works with some extension, while benchmarking against different dosimeters that are available commercially and used extensively in radiotherapy centers.
II. MAtErIALS And MEtHOdS A. the MOSkin detector The MOSkin system is shown in Fig. 1. The MOSkin detector is composed of hermetically sealed MOSFET dye with submicron thickness of the sensitive volume into Kapton pigtail strip with thickness of 0.55 mm using “drop-in” packaging technology(31) (Fig. 1(a)). The thin reproducible polyamide film acts as an electrical connection and buildup for MOSkin, and gives a waterequivalent depth (WED) of approximately 0.07 mm in tissue, making it a suitable dosimeter for skin dose measurement.(24) According to the International Commission on Radiological Protection (ICRP) publication,(21) the most radiosensitive layer of epidermis is located at tissue depth of approximately 0.07 mm. A detailed description of the MOSkin dosimetry system can be found in Kwan et al.(24) and Qi et al.(26) The readout process of MOSkin detector requires measurement of the voltage across the gate of the MOSkin detector under condition of the constant source–drain current that is called the threshold voltage, Vth. The Vth increases with accumulated radiation dose. The readout current corresponds to the thermostable point of the MOSFET to avoid errors associated with thermal instability of the Vth. The sensitivity of the MOSkin detector is defined as the shift of the Vth with the absorption of 1 cGy of radiation dose (Eq. (1)). In this work, the MOSkin measurements were benchmarked against Markus ionization chamber (Markus type 23343 parallel plate ionization chamber; PTW, Freiburg, Germany) and/or GAFCHROMIC EBT2 film (International Specialty Products, Wayne, NJ).
All measurements were carried out three times and the mean ± 1 SD of the readings were reported unless stated otherwise.
(1) where ΔVth is the change of the threshold voltage in unit Volt (V).
Fig. 1. MOSkin system, MOSkin detector (top right), and the schematic diagram of MOSkin detector in (a) face-up and (b) face-down orientation.
B. GAFCHROMIC EBT2 film preparation GAFCHROMIC EBT2 films were cut into sizes of 1.5 × 1.5 cm². They were scanned using a flatbed scanner (Epson 10000XL scanner; Epson America, Inc. Long Beach, CA) 24 hours after irradiation to allow for postirradiation color changes.(32) The films were scanned in a reflection mode, at a resolution of 96 dots per inch (dpi), 48-bits RGB format, and analyzed using ImageJ
1.46r software (National Institute of Health, Bethesda, MD). Care was taken to scan the films at the center of the scanner to avoid scanner-induced nonuniformity. The films were also scanned in the same orientation to avoid film-induced changes in pixel values.(33) Only the red channel was used for analysis. A region of interest (ROI) was selected at the center of the film. A set of standard films was irradiated to establish the calibration curve.
C. detector characterization
C.1 Calibration, linearity, and reproducibility The MOSkin detectors were calibrated under a Varian Clinac 2100 C/D accelerator (Varian Medical System, Palo Alto, CA) using 6 MV photon beam under standard conditions (1.5 cm depth in 30 × 30 × 15 cm³ solid water phantom, 100 cm source–surface distance (SSD), and 10 × 10 cm² field size). Sensitivities of the MOSkin detector have been determined. Linearity measurement of the MOSkin detector was determined for a dose range of 50 cGy to 300 cGy, with an increment of 50 cGy, and the reproducibility was assessed.
In this work, the buildup cap for the Markus ionization chamber was removed in order to position the chamber’s effective measurement closer to the surface. Temperature and pressure correction factor, polarity effect correction factor, and ionization recombination correction factor were taken into account for Markus ionization chamber measurements. Parallel plate ionization chambers (Markus ionization chamber) are known to overrespond due to side scatter from the chamber’s wall.(34-36) In this work, Gerbi and Khan’s correction(36) (Eqs. (2) and (3)) was only applied in the surface dose measurement (Material & Methods section C.4).
(2) (3) where, P′(d,E) is the corrected PDD, P(d,E) is the measured PDD, E is the energy, l is the plate separation (2 mm for Markus PTW 23343), α is constant (5.5), C is the sidewall collector distance (0.35 mm for Markus PTW 23343), IR is the ionization ratio, and d is the depth of the chamber front window below the surface of phantom surface. The calculated ε(d,E) are 10.14 and 6.89 for 6 MV and 10 MV photon beams, respectively.
Except for calibration and dose linearity measurement, all measurements were carried out on the surface of a solid water phantom. Characterization was carried out using a 30 × 30 × 15 cm³ solid water phantom with a 6 MV photon beam, 100 cm SSD, and 10 × 10 cm² field size (Fig. 2), unless stated otherwise. This setup is henceforth called the “standard surface setup”.
Cheung et al.(37) has studied the temperature dependence of this MOSFET-based detector.
They reported that this detector shows a variation of 50 mV over the temperature range from 20°–40°C. This variation is corresponding to about 10 cGy in dose. However, in order to get an accurate reading, the detector should be placed on phantom or patient approximately 60 s before measurement, to allow thermal equilibrium, and the reading are taken whilst the detector remains on the phantom or patient. The same precaution was also taken throughout this work to reduce the effect of temperature dependence of the detector.
Fig. 2. Schematic diagram of the standard setup of characterization of MOSkin detector on the surface of a solid water phantom.
C.2 Source-to-surface distance dependence The MOSkin detector was positioned as per the “standard surface setup”. The response of the MOSkin detector for different distances from the source was measured with SSD, varying from 80 cm to 110 cm with 5 cm increments. One hundred MUs were delivered for each and repeated twice for all measured SSDs. Dose-rate dependence of the MOSkin detector was also evaluated in this section. The dose rate at dmax was calculated.
C.3 Field size dependence The MOSkin detector was set up per “standard surface setup” and irradiated with different field sizes from 1 × 1 cm² to 40 × 40 cm² using 6 MV photon beam.