Development of a Quality Assurance Strategy for Magnetic Resonance Imaging in Radiotherapy

Abstract: Radiotherapy is a common treatment option for malignant tumors. The use of magnetic resonance imaging in radiotherapy is increasing world-wide and sets additional requirements on the image quality compared to diagnostics applications. Thereby, the overall goal of the present project was to develop and implement a structured image quality assurance approach for MRI in radiotherapy. Specific aspects studied was the image quality, spatial accuracy and spin-lattice relaxation time quantification. Measurements using three different phantoms were performed at the positron emission tomography magnetic resonance (PET/MR) scanner at Norrlands Univeristy Hosptial, Umeå. The image quality was evaluated using the American College of Radiology (ACR) phantom and its accompanying tests. ACR phantom measurements were performed once or twice daily throughout the project. The results from the ACR tests were implemented in an automatic analysis program and displayed on a website, accessible from the hospital network. The image quality was found to be constant over time and within acceptance values. Furthermore, the spatial distortion was quantified using a large field of view (FOV) phantom and a spatial analysis program supplied with the phantom. A method to study the reproducibility of the distortion was developed and the distortion in the PET/MR scanner was mapped in different images planes. The image distortion was shown to be reproducible and less than 1 mm inside a volume of 20x20x20 cm3 around the magnetic isocenter, which corresponds to a FOV used in target delineations images. However, outside this volume larger distortions than 2 mm was observed. In order to evaluate the quantification of the spin-lattice relaxation time two methods were evaluated, variable flip angle (VFA) and two point inversion recovery spin echo (IRSE) using the Test object 5 (TO5) phantom. Large variations in the quantification was observed using the VFA method compared to the two point IRSE method when performing phantom measurements. Unfortunately, the two point IRSE method was not adequate when used in head exams in the clinical practice.