Proton therapy (PT) is expected to greatly benefit from the integration with magnetic resonance imaging (MRI). This holds true especially for moving tumors, as the combination allows tumor motion tracking and subsequently a gated treatment or real-time treatment adaptation. At the time of starting the research work as described in this thesis, only one research-grade prototype 0.22 T MRiPT (MR integrated proton therapy) system existed at a static horizontal proton research beamline. The technical feasibility of imaging at that beamline has been presented previously (Schellhammer, 2019). However, a detailed magnetometric study of magnetic field interactions between the MRI scanner and all components of the proton therapy facility was missing so far. Furthermore, to bring the concept of MRiPT towards clinical application, active proton beam delivery seems essential (Oborn et al., 2017). Therefore, the aim of this thesis is to exploratively investigate the feasibility of integrating an MRI scanner with an actively scanned proton beam, focussing on the magnetic field interactions between the MRI and PT systems and their effects on MR image quality. In the first part of this thesis, a study is described which shows the effects of (1) different positions and rotation of the gantry in the nearby treatment room, (2) the operation of the static proton beamline in the research room, and (3) the operation of the treatment room beamline on the B0 field of the in-beam MRI scanner. While the operation of the gantry was found to have negligible effect on the resonance frequency and magnetic field homogeneity of the in-beam MRI scanner, the operation of the two beamlines was found to result in a beam energy-dependent change in resonance frequency on the order of 0.5 μT (20 Hz). This measured change in resonance frequency results in an apparent shift of the MR images. This effect was observed in a previous image quality study during simultaneous imaging and static irradiation performed with the same setup (Gantz et al., 2021; Schellhammer, 2019). It is therefore mandatory to monitor all beamline activities and synchronize the MR image acquisition with the operation of both beamlines in order to guarantee artefact-free MR images and the correct spatial representation of objects in the MR images. Furthermore, a daily drift of the static magnetic field of the MRI scanner was observed and could be correlated to ambient temperature changes, indicating limitations in the heating and the thermal insulation of the permanent magnet material of the MRI scanner. However, this drift can be accounted for by an optimization of the MR frequency calibration prior to each image acquisition. The second part of this thesis presents the combination of the in-beam MRI scanner with an actively scanned proton beam at a Pencil Beam Scanning (PBS) beamline. The investigation focusses on the influences of the magnetic fringe fields of the PT system onto the MR image quality. First, the suitability of the beam-stopper is shown. Moreover, the maximum radiation field of the beamline for operation with the MRI scanner at the beamline is theoretically presented and confirmed by radiochromic film measurements. In order to prevent a direct irradiation of the MRI scanner, it is shown that a reduction of the field size in vertical direction to 20 cm is required, while the full 40 cm field size is applicable in horizontal direction. Furthermore, a beam energy-dependent trapezoidal distortion of the rectangular radiation field induced by the B0 field of the MRI scanner is, for the first time, experimentally quantified at the isocenter of the MRI scanner and confirms previously published computer simulation studies (Oborn et al., 2015). Additionally, a previously unknown proton beam spot rotation is observed for spot positions in the outer corners of the radiation field, with rotations relative to the main axis of up to 22°, which requires future studies to understand the observed effect. Second, the feasibility of simultaneous imaging and dynamic PBS irradiation is investigated, by (1) a magnetometry study and (2) MR image quality experiments during simultaneous PBS irradiation. These measurements reveal that the operation of the horizontal scanning magnet results in a severe loss of image quality in the form of ghosting artefacts along the phase-encoding direction, whereas vertical beam scanning and proton beam energy variation is found to cause no visual degradation of image quality. The origin of the observed ghosting artefacts is unravelled by phase-offsets in the k-space information of the acquired images. A theoretical description of these artefacts is presented, which is capable to explain the experimentally observed image artefacts due to the B0 field perturbations found in the magnetometry study. In order to eliminate the observed artefacts, two concepts for artefact-free imaging during PBS dose delivery are suggested, which include magnetic decoupling of the MRI scanner and PT system, and an online image correction strategy that accounts for the changes in the B0 field caused by the operation of the horizontal scanning magnet. Future studies are crucial to evaluate the feasibility and effectiveness of these approaches. The third part of the thesis tests the hypothesis that a proton beam-induced signal change in MR images, which is indicative of effective proton dose delivery in fluid-filled phantom material, is caused by heat-induced convection (Schellhammer, 2019). It is clearly shown that the inhibition of water flow could fully suppress the beam-induced MRI signal loss that was observed in previous experiments. Furthermore, the introduction of an external flow condition using similar flow velocities as expected during proton irradiation produces similar MRI signal losses. The combination of both results suggests that the observed MRI signal loss is most likely caused by convection and is hence most likely not transferable to solid materials and tissues. However, the method holds potential for the coordinate system co-localization of the MRI scanner and PT system, as well as for verification of the proton beam range during machine quality control. In conclusion, this thesis greatly improves the understanding of the origin and magnitude of perturbations of the static magnetic field of the MRI scanner due to the presence of static and dynamic fringe fields of the beamline and scanning magnets of the PT system. The work shows that these interactions result in a severe loss of image quality during simultaneous MR imaging and active proton beam delivery. Combining the knowledge obtained from magnetometry, imaging and theoretical considerations, solid evidence is provided to understand why this loss of image quality is observed for one scanning direction only. Furthermore, this work shows that the current method used for online MRI-based proton beam visualization is caused by buoyancy-driven convection. These results stimulate further research targeting both non-clinical research solutions and the development of a first prototype MRiPT system for clinical use.:List of Figures vii List of Tables ix List of Abbreviations xi 1 Introduction 1 2 Theoretical background 5 2.1 Proton therapy 5 2.1.1 Physical principle 5 2.1.2 Beam delivery 8 2.2 Magnetic resonance imaging 10 2.2.1 Physical principle of MRI 10 2.2.2 Spatial encoding 12 2.2.3 Basic pulse sequences 13 2.3 Magnetometry for MRI systems 14 3 Magnetometry of the in-beam MRI scanner at the static research beamline 17 3.1 Material and methods 18 3.1.1 Measurement setup 18 3.1.2 Magnetic field camera 19 3.1.3 Magnetic field drift 20 3.1.4 Influence of gantry position and rotation 21 3.1.5 Effect of FBL and GTR beamline magnets 21 3.2 Results 22 3.2.1 Frequency drift and reference measurements 22 3.2.2 Influence of gantry position and rotation 24 3.2.3 Influence of FBL and GTR beamline operation 25 3.3 Discussion 25 4 Combination of the MRI scanner with a horizontal dedicated PBS Beamline 29 4.1 Installation of the MRI scanner at the PBS beamline 29 4.2 Position verification of the beam-stopper 31 4.3 Determination of maximum radiation field size inside the MRI scanner 36 4.4 Discussion 40 5 Magnetic interference and image artefacts during simultaneous imaging and irradiation 41 5.1 Material and methods 41 5.1.1 Magnetometry of external influences on the magnetic field of the MRI scanner 42 5.1.2 Image quality experiments 44 5.1.3 Theory and computer simulation 45 5.2 Results 47 5.2.1 Magnetometry results 47 5.2.2 Image quality experiments 50 5.2.3 Computer simulation 51 5.3 Discussion 52 6 Proton beam visualization by online MR imaging: Unravelling the convection hypothesis 59 6.1 Material and methods 60 6.1.1 Experimental setup 60 6.1.2 MRI sequence design 62 6.1.3 Baseline experiments: Validation of beam energy and current dependency 63 6.1.4 Flow restriction and inhibition 65 6.1.5 External flow measurements 66 6.2 Results 68 6.2.1 Baseline experiments 68 6.2.2 Vertical flow restriction and flow inhibition 71 6.2.3 MRI signal loss by external flow 73 6.3 Discussion 74 7 General discussion and future perspectives 77 7.1 General discussion 77 7.1.1 Magnetometry of the in-beam MRI system 77 7.1.2 Simultaneous MR imaging and active PBS beam delivery 79 7.1.3 MRI-based proton beam visualization 80 7.2 Future perspectives for MRiPT 82 7.2.1 Short-term perspectives 82 7.2.2 Long-term perspectives 83 7.3 Conclusion 87 8 Summary 89 9 Zusammenfassung 93 Bibliography 97 Appendix 109 A Results of film measurements at MR isocenter 109 B Angio TOF MRI pulse sequence parameters 110