Beam model creation and dosimetric validation for preclinical research with X-rays

Abstract

Medical research in the field of cancer therapy is crucial, as cancer is one of the most common causes of death. Radiotherapy is one of the main pillars in cancer therapy and ongoing research aims to improve and expand treatment methods. Pre-clinical in-vivo studies are the final step in the transition from in-vitro cell experiments to clinical application. A prerequisite in small animal research is to reproduce the conditions of radiation therapy (RT) on humans as closely as possible. This indicates that scaling down the entire geometry is essential.In collaboration with the MedAustron Ion Therapy Center (Wiener Neustadt, Austria) and the University of Applied Sciences Wiener Neustadt (Wiener Neustadt, Austria) the Department of Radiation Oncology at the Medical University of Vienna is currently working on establishing the technological basis for high precision image-guided irradiation of small animals with ion beams and kilovoltage X-rays.In this thesis, a beam model for a 200 keV photon beam was developed, which is used as reference irradiation system. The beam model was created in the treatment planning system (TPS) μ-RayStation (RaySearch Laboratories, Stockholm, Sweden), which is specially designed to create dose maps for small animals.In prior work, a measurement setup was developed consisting of an in-house developed beam collimation system, a positioning table and a couch for the anaesthetised small animal, as well as a small field dosimetry phantom (SFDP) for dosimetric measurements. The collimation system consists of a primary collimator (PC) and various exchangeable secondary collimators (SCs) with diameters between 5 mm and 30 mm. Crucial for the accuracy of the beam model was the characterisation of the collimated photon beam. For this purpose, depth dose profiles (DDPs) and lateral dose profiles (LDPs) of the beam were measured using the SFDP with GafchromicTM EBT3 films (Ashland Inc., Wayne, NJ, USA) and a microDiamond detector (PTW-Freiburg, Germany). The field sizes of the SCs were determined with the LDPs taken at the reference point. The 8 mm aperture was chosen as reference field size and the dose rate was measured for this field size at the reference point. The evaluated measurement data was imported into the TPS and irradiation plans were prepared for the validation of the beam model. Targets for the 5 mm, 8 mm, 15 mm and 30 mm diameter SCs were created in the planning module of the TPS and positioned at different depths (3mm, 10mm, 30mm, 50mm). For these targets, treatment plans with a prescribed dose of 100, 200 and500 cGy were created. Hence, the dose distribution and the required irradiation time for the photon beam were calculated by the TPS. To validate the beam model, the deviations between calculated and measured dose were determined. The dose was measured in the SFDP with the microDiamond detector. The best agreement was found for the 8 mm aperture, which was also used as reference field size of the beam model. For this 8 mm SC, in total 10treatment plans were validated with a maximum deviation of under 2%. For the larger SCs, the dose was underestimated by the TPS by up to +8.8% for the30mm aperture and +5.9% for the 15 mm aperture. The largest deviations were found for the 5 mm aperture. In 50mm water equivalent depth (WED) -10.2%dose deviation was measured.Considering the measurement depths, the mean relative deviations were between-0.7% and +2.2%. The highest average deviation of +2.2% was found for the measurements close to the surface in 3 mm WED. The best agreement with ±0.1%mean relative deviation was found for the values in 11 mm WED and 51 mm WED.The determined deviations were in agreement with reports in recent literature about similar studies.The dose deviations for the smallest SC (5 mm) underlined the requirements of accurate positioning. However, the currently used positioning table is limited inits stability, resulting in a decreased level of accuracy. Therefore, an improved version of the table is highly desired to enhance precision and reliability.