Search

This thesis evaluated the application of MRI guidance in different steps of the breast cancer radiotherapy workflow and describes how to handle several challenges for MRI-guided radiotherapy. The application of MRI is most beneficial in the setting of preoperative breast radiotherapy. In this setting, MRI is required for preoperative target definition and the thesis describes consensus guidelines for this purpose. An opportunity for the application of MRI lies in response assessment of changes in the tumor after preoperative radiotherapy. A challenge for MRI-guided radiotherapy is that radiotherapy plans are usually calculated on CT scans and MRI cannot be directly used for dose calculations. This thesis evaluated two different synthetic CT approaches that both resulted in accurate dose calculations for partial breast irradiation. Patient positioning for MRI-guided radiotherapy was also evaluated. A comparison of MRI acquired in prone position and supine position showed that dose to the ipsilateral lung was reduced in prone position. However, appropriate treatment plans for partial breast irradiation on an MR-Linac (a hybrid machine integrating MRI and radiotherapy delivery in a single machine) could be achieved in both positions. The current applications of MRI and the MR-Linac for breast radiotherapy are still limited. This is due to several remaining challenges, but in particular because of the limited value of MRI for the standard treatment in which the breast is irradiated after surgery.

The electrical properties of tissues (permittivity and conductivity) regulate the effects of electromagnetic (EM) fields in the human body. Knowledge of the electrical properties (EPs) permits assessing the safety of human exposure to em fields as generated by telecommunication and medical devices (e.g. mobile phones and magnetic resonance systems). It also allows reliable planning of medical treatments using EM fields. Examples of these treatments are radiofrequency hyperthermia for cancer therapy and EM stimulation for neurological disorders. The EPs are intrinsic tissue characteristics and hold the promise of being endogenous biomarkers, i.e. they could be valuable to discriminate a pathological condition or monitor treatment effects. Magnetic Resonance (MR)-based Electrical Properties Tomography (EPT) is a powerful technique to measure such properties non-invasively. This is possible because the EPs perturb the spatial distribution of the complex B1+ field, i.e. the magnetic field responsible for spin excitation in MR imaging (MRI). MR-EPT requires the acquisition of both amplitude and phase of the complex B1+ field, from which subject-specific permittivity and conductivity maps are reconstructed. Both B1+ acquisition and EP reconstruction can be performed in different ways, as outlined in Chapter 1, and ultimately influence the accuracy and precision of EP maps. Assessing the accuracy and precision of EP maps is a central theme in this thesis. These characteristics determine the validity of EPT as a quantitative mapping tool for clinical applications. This thesis specifically explored different aspects of an MR-EPT experiment that define the accuracy and precision of EPT-based maps: the MR acquisition, the EPT reconstruction and the intended clinical application. First, the impact of acquisition techniques on EPT was quantified: in particular, Chapter 2 focused on B1+ phase mapping methods and their impact on Helmholtz-based EPT (H-EPT) conductivity reconstruction, and Chapter 3 dealt with the influence of three commonly available B1+ amplitude mapping sequences on H-EPT permittivity reconstruction. Then, Chapter 4 presented a new deep learning-based EPT (DL-EPT) approach for conductivity reconstruction in the pelvic region and analyzed its reconstruction performance. Finally, Chapter 5 was extensively dedicated to hyperthermia treatment planning (HTP), the medical application of interest in this thesis. Here, DL-EPT and advanced HTP elements were combined within a single workflow to enable more reliable hyperthermia treatment plans. In conclusion, this thesis implemented technical frameworks on which future research may further build to study the acquisition and reconstruction aspects contributing to the accuracy, precision and clinical applicability of EPT maps. Furthermore, this work paved the way for more reliable, clinically feasible and personalized hyperthermia treatment planning.

MRI is increasingly being used in radiotherapy, because of its superior contrast between the different organs and tumors in comparison to other imaging techniques. For diagnosis, MRI is already standardly used and with the recent clinical availability of hybrid MRI-linac systems, which can acquire MR images during irradiation, the role of MRI is expected to increase even further. There are three stages in which MRI is used, i.e. diagnosis, generation of a treatment plan and treatment guidance. During all three stages, physiological motion (e.g. breathing) is a problem. Due to physiological motion during MRI, a false representation of the anatomy is generated with distortions in the images. As a consequence, these MR images cannot be used for clinical diagnosis or treatment plan generation. During radiotherapy, physiological motion could lead to an underdosage of the tumor and/or overdosage of healthy organs. When motion during MRI and MRI-guided radiotherapy can be measured, the negative effects can be corrected for. For diagnostic scans, the distortions in the images can be reduced. Periodic motion, such as breathing, could be incorporated into the treatment plan by estimating the motion from a 4D-MRI, which describes the anatomy during a full cycle of the periodic motion. During radiotherapy, the treatment could be altered based on the measured motion. In this thesis a novel method to detect motion during MRI and MRI-guided radiotherapy was investigated. This method is based on thermal noise that is measured simultaneous with the MR images and is called the noise navigator. First the physics behind the noise navigator was investigated through numerical electromagnetic simulation on a realistic 4D digital human phantom containing respiratory and cardiac motion. These simulations were validated with measurements on healthy volunteers. Insight was gained on the effect of the setup on the sensitivity of the noise navigator to motion. The second step was to investigate the behavior of the noise navigator during MRI. Practical considerations such as the effect of the number of thermal noise samples and the combination of noise measured by the different channels within an RF receive array were evaluated. Furthermore, a Kalman filter was designed that could enable prospective motion correction based on the noise navigator. Next, the feasibility of respiratory-correlated 4D-MRI generation based on the noise navigator was demonstrated for cartesian and radial readout strategies. These 4D-MR images were compared to 4D-MRI generated with conventional methods. The noise navigator would be valuable for this application as there are at the moment no comparable methods for cartesian acquisitions. Finally, the detection a variety of motion types with the noise navigator for three different anatomical applications of MRI-guided radiotherapy were investigated. It was feasible to simultaneously detection bulk body movement and respiration in the torso, swallowing for head-and-neck, and cardiac activity. Although the latter requires more research.

RF transmit signals are an inherent part of MRI that is crucial for spin excitation. Current developments in MRI are moving towards higher magnetic field strengths and more transmit coils leading to more inhomogeneous RF distributions. This increases the possibility of SAR hotspots, creating higher risks of local heating. In order to cope with these developments and still perform reliable safety assessments the MRI safety community is discussing to revise the guidelines on RF induced tissue heating. Instead of restricting the scanners on a derived measure, i.e. SAR, the scanners should be limited by the guidelines that are direct measures for tissue damage such as absolute temperature. In this debate the translation from SAR to temperature estimations is crucial. In order to make this translation thermal modeling is required. In this thesis MR-based methods were developed in order to measure the influence of RF exposure in human subjects, both in temperature and perfusion. Those results could subsequently be used to test whether current thermal modeling was able to predict those effects for a given subject. In vivo temperature measurements in the calf resulted in precise representations of the heating that could be used as a reliable gold standard. Furthermore, perfusion increases over the complete leg due to this local temperature increase could be observed. Although the simulations matched with the experiments in the EM regime, the accuracy of the estimated temperature distributions was not sufficient for a safety assessment. This showed that more work on thermal modeling is required before MRI scanners can be restricted on measures based on temperature. Therefore it would be unwise to immediately discard all SAR related restrictions. Local SAR still proved to be a very useful restrictive measure since it can be quickly determined by simulations and can more easily be verified. Correct thermal modeling proved to rely on inputs that are not easily measured such as subject specific thermoregulation. Although the first steps in this direction were made in this thesis, obtaining a proper estimate of thermal behavior over the whole human body is a highly challenging pursuit.

Radiotherapy cancer treatment uses ionizing radiation to destroy malignant cells. However, radiation also affects healthy tissue surrounding the tumor, which may cause unwanted side effects that can severely impact the quality of life of cancer patients. The aim of radiotherapy is to destroy malignant cells, while minimizing damage to healthy normal tissues. To successfully deliver the radiation dose to the tumor while reducing the dose to the surrounding healthy tissues, it is necessary that the tumor is accurately defined. Three-dimensional medical imaging techniques are an essential tool for tumor delineation. In head-and-neck radiotherapy, tumor delineation is most frequently performed using computed-tomography. The visibility of the tumor can be improved with the additional information provided by other imaging techniques, such as magnetic resonance imaging and positron emission tomography. However, in head-and-neck cancer, the precision of tumor delineation is still low and one of the largest sources of uncertainty. This suggests that there is still need for better interpretation of imaging. Pathology is considered the gold standard for interpreting and validating imaging techniques. To validate imaging techniques with pathology, imaging scans performed before surgery should be compared with pathology data obtained from the resected surgical specimen. The aim of this thesis is to validate advanced imaging techniques with pathology for tumor delineation and characterization in laryngeal and hypopharyngeal cancer radiotherapy.

First, we present two offline position verification methods which can be used in radiotherapy for detecting the position of the bony anatomy of a patient automatically with portal imaging, even if every single portal image of each segment of an (IMRT) treatment beam contains insufficient matching information. Additional position verification fields will no longer be necessary, which speeds up the treatment and reduces the total dose to the patient. Second, a tool is described which enhances the way tumors can be delineated by using multiple imaging modalities. This tool is especially useful when multiple MRI sequences are available as well as the standard planning CT. We also developed a marker which is visible on MRI and on EPID. A separate CT for detecting the markers is no longer needed. The gold marker with steel core can be detected on various MRI sequences, reducing the overall systematic radiotherapy treatment error. The MRI linear accelerator (MRL) facilitates continuous patient anatomy updates regarding translations, rotations and deformations of targets and OAR during a course of radiotherapy. Accounting for this information demands high speed, online IMRT re-optimization. Therefore, we developed a fast IMRT optimization system which combines a GPU based Monte-Carlo dose calculation engine for online beamlet generation (GPUMCD) and a fast inverse dose optimization algorithm (FIDO). We show that for the presented cases the beamlets generation and optimization routines are fast enough for online IMRT planning. Furthermore, there is no influence of the magnetic field on plan quality and complexity, and equal optimization constraints at 0T and 1.5T lead to almost identical dose distributions. One of the most significant effects of the transverse magnetic field on the dose distribution occur around air cavities: the electron return effect (ERE). We investigated the effects of non-stationary spherical air cavities on IMRT dose delivery in 0.35T and 1.5T transverse magnetic fields by using Monte Carlo simulations. Our observations show the intrinsic ERE compensation by equidistant and opposing beam configurations for moving spherical air cavities within the target area. IMRT gives some additional compensation, but only in the case of correct positioning of the air cavity according to the IMRT compensation. For air cavities appearing or disappearing during a fraction this correct positioning is absent and gating or plan re-optimization should be used. Finally, we introduce an online 'virtual couch shift' (VCS): we translate and/or rotate the pre-treatment dose distribution to compensate for the changes in patient anatomy and generate a new plan which delivers the transformed dose distribution automatically. The VCS is the first step towards compensating all anatomical changes (translation, rotations, and deformations) by online re-optimization of the IMRT dose distribution.

The treatment of advanced cervical cancer patients consists of chemo-radiation and brachytherapy. After decades of conventional treatment, 3D imaging (CT and/or MRI) was introduced into the brachytherapy treatment procedure. MRI has been shown to be superior to radiography and CT for accurate tumour differentiation, and is therefore recommended for treatment planning. MRI-guided brachytherapy allows tailoring the dose to the target volume, resulting in high tumour doses. Furthermore, MRI information revealed the limitations of the traditionally used intracavitary applicators, and resulted in the development of the MRI compatible Utrecht applicator. The Utrecht applicator is a tandem-ovoid applicator of which the ovoids are used as a template for interstitial needle placement. Insertion of the needles proved to be feasible and safe and provided a better coverage of the target, while sparing the surrounding organs. Different centres deal differently with the use of additional needles. A treatment comparison study showed that some centres tailor the dose to the target volume, while others are more conservative in maintaining their traditional peer-shaped dose distribution. Interestingly, another finding of this study was that MRI-guided brachytherapy treatment planning can result in similar dose parameters while the dose distribution varies. The question whether high dose areas will lead to necroses and more severe morbidity can only be answered when during meaningful follow up morbidity is scored. An outcome and morbidity study of the first patients treated with chemo-radiation and MRI-guided brachytherapy in the UMC Utrecht showed excellent outcome, especially with respect to local control and acceptable treatment related morbidity. Despite respecting the dose constraints for the surrounding organs, vaginal and gastro-intestinal morbidity occurred in 5 and 10% respectively. Unfortunately, no association between severe morbidity and dose parameters could be found. The unique setting of a 1.5 Tesla MRI scanner integrated in the brachytherapy theatre of the UMC Utrecht allowed additional MRI scanning during the course of brachytherapy. For pulsed dose rate treated patients (delivered in about 30 hours) re-imaging was performed ones during treatment and revealed an increase in rectum dose systemically over time. This might explain that no dose response relationship for gastro-intestinal morbidity could be found. A systematic increase in dose was not seen in high dose rate (HDR) treated patients. For HDR treatments, intra-fractional dose differences for the organs were generally small. However, incidental high rectal dose deviations were found in individual patients, due to rectal volume changes. Re-imaging before irradiation is valuable, at least for treatment plans where dose parameters meet the organ dose constraints. If necessary, small interventions can help to stabilize the anatomical position before dose delivery. Overall, the impact of organ movement on the target dose was small. Chemo-radiation and MRI-guided brachytherapy in the UMC Utrecht results in an excellent 3-year local control rate of 93%. The 3-year overall survival and progression free survival rates were 65% and 71%, respectively. When comparing with our conventional treatment approach, all outcome parameters were improved. The high local control rate gives us the opportunity to develop treatment strategies to improve regional and distant control.