| Popis: |
Positron emission tomography (PET) is a powerful tool in nuclear medicine allowing the visualization and quantification of metabolic processes down to the molecular level. As a tracer-based imaging modality, PET images lack anatomic information required for co-registration and localization, thus another imaging system is required to contribute the missing information. A highly regarded candidate is magnetic resonance imaging (MRI) which offers besides anatomic images with great soft-tissue contrast also a wide range of other contrast mechanisms allowing the extraction of e.g. functional information. Hence, combined PET/MRI became an interesting hybrid modality. The integration of those two systems into a hybrid devices allowing for simultaneous data acquisition is however a challenging endeavor: as the MRI system makes use of strong, static (the magnetic B0 field) as well as fast switching electromagnetic (EM) fields (the gradient fields and the radiofrequency (RF) field), it provides a harsh environment for any electronic device (the PET scanner) designated to be operated inside the bore of the MRI scanner. On the other hand, MRI is known to be a low-sensitive imaging modality whose image acquisition principle and image quality relies on strict quality criteria with respect to the applied EM fields (homogeneity, linearity etc.). Therefore, the presence of a PET scanner might severely influence the imaging process of the MRI system. Hence, the MR compatibility of a PET scanner is serious issue and the assessment of interference phenomena is important to evaluate the limitations on the imaging process and to understand the problems occurring.In this thesis, the MR compatibility of preclinical PET insert, the Hyperion-II-D detector, is assessed. The insert is designed to be operated inside a clinical 3T MRI scanner and makes use of a digital implementation of Silicon Photomultipliers as photo detectors which were firstly presented by Philips Digital Photon Counting. In contrast to other PET/MRI research prototypes, most of the readout electronics are operated inside the bore of the MRI system, thus potentially intensifying the MR compatibility issue. As no standardized evaluation protocols exist for PET/MRI devices, a set of test protocols, which allows a reliable assessment of the MR compatibility, is developed in this thesis, whereby the focus lies on the application of technical, synthetic test protocols which allow the differentiation of interference phenomena between the PET insert and the individual subsystems of the MRI scanner. Overall, an acceptable level of interference was observed: although the PET scanner influences the MRI acquisition noticeably, the degradation effects can be considered as acceptable for morphological scan as well as for more advanced acquisition schemes. On the PET side, a degradation of the energy and timing resolution was observed when highly demanding MRI test sequences, which produce maximum demanding gradient fields, are applied. However, this effect only occurs for extreme scenario and play a subordinate role for normal imaging sequences.Besides the MR compatibility assessment, a B0 optimization approach on PET detector level is proposed: as any material brought inside the bore of the MRI scanner distorts the B0 field, the B0 distortion can be regarded as an inherent MR compatibility issue which is counteracted by avoiding materials for the PET development which have a high magnetic susceptibility. As this design paradigm limits the design flexibility, the application of additional (highly magnetic) material (passive shimming) or conductor loop configurations (active shimming) on PET module level is proposed in this thesis. The resulting B0 distortions originating from the PET module and the additional materials superimpose yielding a homogenized B0 field distribution if applied correctly. A simulation and optimization toolkit is developed in thesis and first proof-of-principle experiments are conducted on basis of single PET modules. |
