Abstract
Non-invasive image-guided interventions such as high-intensity focused ultrasound (HIFU) or external beam radiotherapy (EBRT) play next to invasive surgery and chemotherapy an increasingly important role as one of the cornerstones in contemporary cancer care. Both techniques aim to destroy tumor cells by directing high energy to the tumor while leaving
... read more
surrounding tissue intact. Their non-invasiveness is the key strength of these techniques, providing potential treatments for patient groups that are not eligible for surgery or even providing an alternative to surgical resection. Due to anatomical and physiological motion, however, therapy delivery via these two non-invasive modalities becomes challenging, especially for organs in the thorax and abdomen. In case the displacements of the target pathology are not taken into consideration, the therapeutic energy may end-up being diverted from its intended delivery location, with two possible consequences: 1) A sub-lethal amount of therapeutic energy is delivered to the pathology, increasing the risk of under-treatment and implicitly of disease recurrence; 2) In case motion effects are severe, healthy tissues may move into the beam-path, resulting in potential collateral damage to critical structures. Therefore, motion compensation schemes are crucial for achieving the therapeutic end-point of HIFU and EBRT interventions in moving organs, while at the same time maximizing healthy tissue sparing. Medical image-based motion compensation is particularly attractive since it provides a non-invasive means for therapy planning, guidance and response monitoring. Moreover, the fact that an increasing number of therapeutic systems are fitted with on-board medical imaging devices, further facilitate the widespread usage of such approaches. A large selection of image-based motion correction solutions for HIFU and EBRT are already available in the literature, however, the development and adoption of a method which definitively addresses this issue is particularly difficult. This is due to a variability in physiological properties across different pathologies and anatomies, the large degree of freedom in selecting imaging devices and/or image acquisition parameters and a high variability in inter-patient anatomical motion characteristics. For this reason, adapted motion estimation/compensation schemes are necessary, optimized for a particular application/scenario. Therefore, if a standard motion compensation strategy is to be translated into clinics, a tendency towards specialized multi-component frameworks, addressing a broad category of displacements and encompassing the entire work-flow of the intervention is desirable. This thesis makes a step towards such a strategy, providing an ensemble of novel motion estimation and compensation solutions, which when fused together, become a large-scale framework encompassing the therapeutic work-flow of HIFU and EBRT interventions.
show less
