dc.description.abstract
In preclinical radiation oncology settings, the use of kilovoltage X-rays is considered a standard irradiation method. Simultaneously, with growing interest in high-precision irradiation techniques, conventional X-ray units often lack the equipment for small field applications. Therefore, not only a precise evaluation and characterization of the X-ray unit itself is necessary, but also the development of a suitable setup to meet these requirements. The first part of this work focuses on analyzing the dosimetric characteristics of an oil-cooled X-ray unit (Y.Solution Maxishot). This X-ray tube uses tungsten as target material at a target angle of 20 ° and is operated at the maximum working voltage of 200 kV and a current of 20 mA. The analysis involved examining the dose rate stability over the long term and during the X-rays warm-up phase, as well as the beam quality, homogeneity of the radiation field, depth dose distributions, and the behavior of the detectors used. Various filter combinations were utilized, including 3 mm beryllium, 3 mm aluminum, and 0.5 mm copper. The aluminum filter and copper plate can be removed as required. The dose profiles were determined using the Semiflex (PTW31013) and the microDiamond (PTW60019) detectors. Polymethyl methacrylate (PMMA) phantoms were used to mimic the biological tissue of small animals, as these have proven to be an effective substitute for water. The long-term quality assurance (QA) evaluation, measured over a period of one year using the Semiflex detector, showed an annual decrease of 1.5 mGy/min or 0.16 % per month. During the ramp-up phase of the X-ray unit, the dose rate remained constant for exposure times exceeding 2 s, with a quality factor of R2 close to unity. This indicates that no correction factor is necessary for preclinical studies. To measure the quality of the beam, it is important to consider the working voltage of the X-ray tube and the material thickness at which the dose is halved from its maximum (HVL). This resulted in measured values of 13.35 mm Al and 1.09 mm Cu, after filtering through 3 mm Be, 3 mm Al, and 0.5 mm Cu. The deviation from simulated values using SpekPy ranged from 3.7 % to 6.2 %. The microDiamond detector exhibited an angular dependency of ±3 % for orientation perpendicular, and ±1 % for orientation parallel to the beam axis. A difference in the response between both detectors of up to 4.5 % was observed in measurements with small RW3 cubes ((3×3×3) cm3 and (5×5×5) cm3), indicating different sensitivities to backscattered radiation from the material behind the detector. To determine beam homogeneity in a (210×200) mm2 field, detectors were placed in a reference depth of 11 mm PMMA with different filtration combinations. With only 3 mm Be filtration, clear inhomogeneity was evident in the horizontal positioning, with a drop in dose of over 10 % at the edges of the field (±100 mm). However, with the addition of 3 mm Al and 0.5 mm Cu filters, the homogeneity could be improved, with deviations of only 2 % in the central field area (±50 mm). For depth dose profiles in PMMA, a dose gradient ranging from 11.3 %/cm (with only 3 mm Be filter) to 11.7 %/cm (with the Be, Al, and Cu filters) was achieved with the Semiflex detector. In contrast, the microDiamond detector exhibited a less steep dose distribution, ranging from 9.3 % to 9.6 %. Lastly, the dose distribution within a setup for the whole-body irradiation of up to ten mice was tested. A homogeneity of up to 3 % was visible in the central area (columns 2–4) within a 5×2 field grid. It was noticeable that the two outer columns were not suitable for irradiation, given that dose deviations of up to 11.6 % were recorded.A modular brass collimator system with eight interchangeable secondary collimators (SC), with diameters ranging from 5 mm to 30 mm, was developed for precise irradiation of small target volumes. The system also includes a primary collimator with a 46 mm aperture and a 114 mm tube that can be fixed directly to the existing X-ray unit with screws. In addition, a phantom holder (SFDP) was designed that can hold up to 150 RW3 plates, each with dimensions of (60×60×1) mm3. The dosimetric evaluation of the new system was performed using EBT3 films and the microDiamond detector, with a reference depth of 20 mm RW3 and a reference collimator aperture of 8 mm. Values were measured at a source-to-detector distance (SSD) of 291 mm and a source-to-axis distance (SAD) of 311 mm. Based on the measured horizontal dose distribution at the reference depth, a correlation was determined between the aperture and the beam parameters. The deviation from measured full width at half maximum (FWHM) and the aperture was (1.31 ± 0.02), the flatness of the lateral dose profile varied from (5.08 ± 0.15) % (15 mm aperture) to (8.59 ± 0.04) % (30 mm aperture), the symmetry of the profile remained below 4 %, and the penumbra value varied from 0.73 mm to 2.27 mm for larger apertures. The maximum transmission through the tube–PC connection was at 5.17 %, and this should be analyzed in more detail for future work. The measured output factors ranged from 0.9 (5 mm) to 1.31 (30 mm), when normalized to the reference collimator value at the reference depth. It was also noticeable that, for short irradiation times of less than 40 s, a decrease in dose rate of up to 4.8 % was observed, which must be considered for preclinical applications. The dose rate subsequently stabilized at (1.593 ± 0.018) Gy/min when measuring the reference collimator at the reference depth.The acquired beam data were the input data for the beam model of the dedicated treatment planning system (μ-Raystation), based on measured percentage depth dose (PDD) profiles, lateral dose profiles (LDPs), output factors, and collimator and measurement depth geometries. The treatment planning system simulation was in good accordance with the measured PDDs at a depth of 20 mm with a discrepancy of less than 0.01 %, but there was a tendency to overestimate the dose at shallow depths and for small apertures (by up to (9.8 ± 0.4) %). An exemplary treatment plan was generated based on a μ-CT dataset of a BALB/cJRj mouse using this beam model. A vertical beam and a target structure in the central brain region with diameter of 5 mm were utilized with the 5 mm aperture. The target received an average dose of 2 Gy with complete coverage (V95=100 %) and a homogeneity index of 21 %. Due to the high absorption, a D_(2%) of 6.06 Gy and an average dose of 0.91 Gy in the total volume of the skull were achieved. This is a completely expected effect when using kilovoltage X-rays, due to the photoelectric effect, which leads to significantly higher absorption in bone structures. A control point located below the skull, in the region of the trachea, received a D_(2%) of 2.08 Gy. The oral cavity was explicitly omitted from the beam path to reduce side effects.
en
