Universal parallel transmission pulse design for the human brain and spinal cord MRI at 9.4T

Magnetic resonance imaging is a powerful, non-invasive technology to acquire anatomical images from the human body. Operating at a magnetic field strength of 7T or higher (i.e. ultrahigh field (UHF)) provides a higher signal-to-noise ratio, facilitates higher spatial resolutions, and potentially improves diagnostic sensitivity and specificity compared to clinical field strength, such as 1.5T or 3T. Unfortunately, UHF is accompanied with technical hurdles, from which the most problematic is the inhomogeneity in the radiofrequency transmit field. That can lead to spatially varying flip angles and, thus, to signal dropouts, local brightening or spatially altering imaging contrast. The most flexible approach to address this issue is the parallel transmission (pTx) technique, which itself has the disadvantage of a lengthy calibration procedure. To overcome the calibration procedure the ‘universal pTx pulse’ (UP) concept was introduced. It is a radiofrequency pulse design concept that relies on a pre-collected design database. The resulting pulses then work on a wide cohort of subjects without recalibration. As a first step, in this PhD project the advantages of imaging the human spinal cord at UHF were exploited. It was possible to acquire the first images from the human spinal cord at an ultrahigh in-plane resolution of 0.15x0.15mm2 at 9.4T. The images showed the tiny structures of the spinal cord in great detail. The signal-to-noise ratio and T2 *- times in the human spinal cord at 9.4T were presented. Furthermore, in this thesis the UP concept was further developed, in order to use UHF and the pTx technique more widely. While UPs were originally introduced for whole-brain or slice selective excitation, in this work a feasibility study for UPs for local excitation in the human brain (i.e. exciting only specific regions of the brain, while others should experience no excitation) was performed. UPs that locally excite the visual cortex area were calculated. The underlying transmit k-space trajectory for these radiofrequency pulses were ‘spiral’ trajectories. These local excitation UPs were successfully tested in vivo on nondatabase subjects at 9.4T. In a next step, the UP performance was further improved by optimizing the underlying transmit k-space trajectory to match the excitation target. The trajectory optimization and the UP design algorithms have been implemented into an open source software package (called OTUP) and demonstrated using simulations and in vivo experiments at 9.4T. The code was tested for three different target excitation pattern with varying complexity