Abstract
In recent years a paradigm shift in therapy evaluations for oncology is apparent. Focusing on the morphological changes with the variety of imaging modalities is not sufficient to assess therapy efficacy during treatment.
In the first part of this thesis the changes of relaxation times were quantified in ex vivo
... read more
porcine recta during formaldehyde fixation and in human specimen. Chapter 2 concludes that T1 of the circular muscle, mucosa and submucosa, and T2 of the circular muscle decreased. Consequently, fixation, prior to imaging, progressively decreases image contrast. The human specimens were therefore imaged during the first few hours of fixation, in search for a suitable diagnostic marker. However, no conclusive parameter resulted as a significant diagnostic parameter for therapy evaluation.
First step towards bringing metabolic imaging to patients is good magnetic field (B0) uniformity. Improvements of the B0-field homogeneity by simulating a local array of shim coils on liver MRI during different respiratory states are investigated in chapter 3. From simulations it is apparent that the increased degree of freedom improves homogeneity to a greater extent (10%) than using conventional shim hardware. Shimming B0 dependent on respiratory states reduces the effects of the respiratory motion with improvements ranging from 21 to 44 %.
Chapter 4 facilitates the next step by integrating a 31P whole-body birdcage coil in 7T MR system, to enable 31P MRS over large fields of views. The design allows use of rectangular RF-pulses which decreases overall SAR compared to adiabatic RF-pulses while maintaining flip-angle homogeneity. It permits exchange in SNR for an increase in resolution and/or MR-weighting in 31P MRSI. The number of RF-pulses/dt can be increased as is demonstrated by fast 3D CSI of the liver with a short TR and low flip-angle excitation to acquire optimal SNR/dt. Furthermore, T2 of PCr in the gluteal muscle was quantified using a MESING protocol, something previously not possible large volumes.
Interpretation of accompanied data required development of a software tool, CSIgui as explained in chapter 5. The large data sets have a high number of dimensionalities and are therefore not handled well by current software solutions. The tool is compatible with conventional MRS and raw file types. Data can be processed using a variety of MR algorithms and saved for analysis.
Chapter 6 focuses on the feasibility of acquiring 3D CSI of tumor tissue in four lung carcinoma patients using the 31P whole-body birdcage coil in combination with a dual-Rx coil or a Rx array setup. Simulations proved the loss in SNR/dt with imperfect flip-angles is minimal using a short TR low flip angle protocol, eliminating a time-consuming calibration process. However, use of the receiver array favored SNR over a larger FOV compared to the dual coil setup. Detected metabolites were ATP and Pi, but also the phospholipid metabolites including the individual PME, PE and PC, and PDE, GPC and GPE which are not detectable at 3T. 31P MRSI of lung tumors at 7T is feasible and warrants further research as a non-invasive method for therapy evaluation.
show less
