An abdominal phantom with anthropomorphic organ motion and multimodal imaging contrast for magnetic-resonance-guided radiotherapy

The continuous progress in the field of radiation therapy has led to significant improvements in the diagnosis and treatment of cancer, resulting in enhanced quality of life and increased life expectancy. The development of the MR-Linac marked a paradigm shift in radiation therapy. This advancement enables real-time visualization of tumors during radiation therapy using magnetic resonance imaging. Consequently, treatment plans can be adjusted to account for changes in tumor size between sessions, such as tumor shrinkage, and to incorporate tumor movements during each radiation session, for example, due to breathing. This precision allows for the delivery of a higher radiation dose directly to the target volume while minimizing radiation exposure to nearby organs. The aim of this work was to develop an anthropomorphic abdominal phantom that meets several requirements: reproducible breathing motions with induced organ motions in a composite, realistic image contrast in both magnetic resonance imaging and computed tomography, anthropomorphically shaped organ models, and an MRI-compatible motion control unit. In this thesis, an innovative anthropomorphic abdominal phantom for medical imaging and radiation therapy applications was developed. Through a series of experiments and analyses, the capabilities and usefulness of the phantom were rigorously evaluated. The organ models used in the experiments demonstrate remarkable accuracy in replicating the relaxation times and Hounsfield Units of real human organs. This validation underscores the suitability of the phantom for medical imaging research, with the results showing close agreement with reference values without significant differences. Comparisons between the phantom and patient/volunteer data showed good agreement in simulating respiration-induced organ motions in a composite during various breathing patterns (shallow, free, and deep breathing), anatomical shapes, image contrast, and radiological characteristics. Furthermore, the analysis of organ motion under different breathing patterns highlights the phantom's ability to simulate human organ movements, emphasizing the importance of considering organ motions in treatment planning and imaging procedures. In summary, this work demonstrated that the developed phantom effectively simulates various respiratory movements and corresponding organ motions within a composite structure. Additionally, compared to volunteer data, the phantom exhibited comparable image contrast in magnetic resonance imaging and computed tomography imaging, and stability of image contrast over a period of more than 400 days was demonstrated. Moreover, the phantom proved suitable for an end-to-end test, encompassing the entire radiation therapy process from imaging and radiation planning to dose calculation and delivery. This included the insertion of dosimetric EBT3 films into the liver tumor model. An important outcome was that the phantom's liver tumor model was successfully detected by the MR-Linac and radiation was stopped as soon as the tumor moved outside the target volume due to breathing motion. Ultimately, a dose of 5.3 ± 0.42 Gy was calculated within the tumor model, which demonstrates excellent alignment with the planned dose of 5 Gy, considering the minimal deviation.