Degree Name

Doctor of Philosophy


Centre for Medical Radiation Physics


[Background and Purpose] In-vivo, skin, and interface dosimetry in modern radiation therapy is an important issue that demands a reliable radiation dosimeter capable of measuring skin doses. The Metal Oxide Semiconductor Field Effect Transistor “MOSFET” dosimeter became popular as a radiation dosimeter in radiation therapy and one of its recently designed versions for skin dosimetry, called MOSkin, has become very popular. Unfortunately MOSFET has a limited lifespan due to a saturated build up of positive charges in its sensitive dosimetric volume - gate oxide, which is mainly located near the Silicon- Silicon dioxide (Si-SiO2) interface. This built up charge reduces its sensitivity, linearity, and stability. The aims of this thesis are an investigation of MOSkin™, with the possibility of annealing the MOSFET dosimeter designed by the Centre for Medical Radiation Physics, and being able to re-use it again after multiple cycles of irradiation and annealing. Virgin and annealed MOSkin detectors have been reused in addition to an Attix chamber, a Monte Carlo simulation, and EBT2 films to verify a newly developed radiotherapy treatment planning algorithm (TPA), the Acuros® XB (Varian, Palo Alto, CA). This verification covered dose predictions in the areas with strong electronic disequilibrium such as the build-up region and interfaces inside the phantom such as tissue – air, tissue – lung, and tissue – steel. The current clinical treatment planning algorithm (TPA), the Anisotropic Analytical Algorithm AAA (Varian, Palo Alto, CA) was verified experimentally in comparison to the Acuros® XB.

[Materials and Methods] Groups of MOSkin detectors were irradiated with different doses on a 6 MV X-ray beam from a medical linear accelerator (LINAC) and then annealed using different annealing methods. Ultraviolet (UV) light with 4.92 eV mono-energetic photons from a filtered Mercury vapour lamp was used for UV annealing, while isothermal annealing was carried out in the furnace heated to a temperature of 1500C. Direct electric current (DC) was also used to anneal MOSkin with a current ranging from 5 mA to 15 mA, and pulsed electric current was used for current annealing with a current ranging 50- 300mA per pulse. The annealing time required to recover the threshold voltage (Vth) varied for each type of annealing, so the parameters influent on this were studied. The annealed MOSkin detectors were irradiated again to check their sensitivities, linearity, and signal stability. The cycles of annealing and irradiation for each group of MOSkin were repeated many times to investigate the effectiveness of each annealing method. To verify and benchmark the treatment planning algorithms, MOSkin and Attix ionising chambers were used to verify the dose calculated by the Acuros® XB, and the AAA at the build-up region for 6 MV and 18 MV X-ray photon beams with field sizes of 4 cm x 4 cm, 10 cm x 10 cm, and 40 cm x 40 cm, under a normal incidence X-ray beam from linear accelerator (LINAC). To check the performance of treatment planning systems (TPSs) for dose calculations under an oblique beam, experiments with a 45˚ beam incidence and 10 cm x 10 cm field size with 6 MV and 18 MV X-ray photons beams were carried out.

For interface dosimetry, MOSkin and EBT2 films were used to measure doses near the air, steel, and lung interfaces, as well as using Monte Carlo simulations with 6 MV and 18 MV photon beams and 10 cm x 10 cm field sizes, for all but the lung cases, with a small 3 cm x 3 cm field size.

[Results] The annealing methods that were investigated revealed different clinical outcomes. UV light was able to recover the threshold voltage but not the sensitivity with essential MOSkin Vth instability (within ±28 mV). Isothermal annealing was a good alternative in that it fully recovered the sensitivity and threshold voltage with an acceptable signal instability (±4 mV). DC annealing was shown to be much better than UV and isothermal annealing provided full recovery of sensitivity, threshold voltage linearity dose response, and excellent stability within ±1 mV. Pulsed current annealing was the best annealing method of all, with full recovery of the sensitivity, threshold voltage dose response linearity, and excellent stability ±1 mV.

Dosimetry in the build-up region resulted in an excellent agreement between MOSkin and Attix (within ±2%), although Acuros® XB was much better than AAA in all the setups used. For interface dosimetry, Acuros® XB performed better than AAA near the interfaces, although it had some slight shift in depth dose distribution within 2 mm of the proximal and distal interfaces, for all cases except the lung. With the lung case, Acuros® XB and AAA performed satisfactorily and agreed with the Monte Carlo simulated doses within ±5%.

[Conclusions] Direct current annealing (DC) and pulsed current annealing are the best annealing techniques for a MOSkin dosimeter. This will have potential benefits for MOSFET applications in medicine and space dosimetry, especially for departments with limited budgets, and in developing countries.

The verification of Acuros® XB showed that this TPA performed much better than previous TPA and AAA, and it was very close to the MC simulation and experimental measurements. As an application for the annealing procedure, pulsed current annealing was used with build-up measurements. The total doses during the experiments for build-up and interfaces dosimetry were 230 and 185 Gy, respectively. It has been shown that two periodically annealed MOSkins detectors were enough for all the build-up measurements, while for doses on interface dosimetry measurements, the 11 virgin MOSkins that were used clearly showed the benefits of the annealing technique developed for MOSkin dosimeters for clinical applications.