Magnetic induction tomography spectroscopy (MITS) is a contactless, noninvasive near-field imaging modality aiming at the reconstruction of the passive electromagnetic properties of different materials or media. If any material or medium should be a biological one, such as living tissue, then speaking of biomedical MITS, the core theme of this thesis. It is much to be hoped to find MITS in some routine medical applications in the future. However, answering such hopes certainly involves a great deal of arduous research work because living tissues are necessarily expected to be under investigation, and dealing with such complex biological structures as which is definitely most challenging. Therefore, the different problematic issues regarding the measuring and imaging processes have to be comprehensively analyzed and carefully tackled in order to hopefully ultimately activate biomedical MITS in routine clinical use. Furthermore, it is of prime importance to address these issues experimentally rather than only by simulation because many of the problems to be confronted with during the technical implementation of the MITS modality are unpredictable, extremely difficult to simulate or cannot be simulated, the reason why this work is completely based on experimental research. A probably most appropriate starting point in this regard would be the MITS tomograph itself being responsible for the measuring and imaging processes, the reason why this study is completely concerned with tomograph optimization in biomedical MITS. For these purposes, the proposed general optimization approaches were applied to a prototype MITS tomograph. The performed optimization covered the tomograph's hardware, software and measurement environments, and is naturally not to be considered as to be specific to this prototype tomograph, but to any MITS tomograph in general to be optimized to meet the technical requirements for biomedical applications. As regards hardware environment, the optimization essentially concerned its core module, namely, the transceiver system, and hence the relevant subsystems: mechanics, electrics and electronics. In this respect, a novel dual-plane elliptical transceiver system was constructed. As its geometric layout approximates the shape of the human body, especially the truncal regions, it is particularly appropriate for thoracic, abdominal and pelvic applications. As regards software environment, the optimization essentially concerned its core module, namely, the imaging program, and hence the relevant subprograms: data postpro-cessing, sensitivity calculation and image reconstruction. In this respect, a novel hybrid dynamic-parametric imaging concept was developed. The addressed application example was hybrid state-frequency-differential imaging wherein simultaneous local and spectroscopic conductivity changes in the biological target medium were reconstructed together into one and the same image, gained from one and the same measurement. As regards measurement environment, the optimization essentially concerned the detection of unintentional patient movements during measurement and the elimination of their consequences prior to image reconstruction. In this respect, a novel detection and elimination technique (D&E) was originated. It allows detecting and eliminating any motion-induced signal errors whatsoever, whether those caused by the patient, the transceiver system or both together, hence, eliminating any motion artefacts in the images to be reconstructed without any need to repeat the time-consuming measurement and imaging processes.
|Betreuer/-in / Berater/-in|
|Publikationsstatus||Veröffentlicht - 2020|