Simulation, Design, and Implementation of Approaches to RF-Encoded (TRASE) MRI
Date
2022-05-16
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
ORCID
0000-0003-4292-9830
Type
Thesis
Degree Level
Doctoral
Abstract
Magnetic Resonance Imaging (MRI) has become a powerful medical imaging tool in the last four-five decades. Technological advancements in MRI have largely focused on increasing the field strength - to achieve higher signal to noise ratio (SNR) - but there is a strong motivation behind developing low-field MRIs. Low-field MRIs are cheaper, smaller, lighter and lower-power devices in comparison to their standard clinical counterparts. Therefore, low-field MRIs are good candidates for developing portable MRIs for point-of-care (PoC) applications. One prominent need for such portable and low-field MRIs is in space missions, for studying and monitoring astronaut health. TRansmit Array Spatial Encoding (TRASE) is a novel MRI technique in which the spatial encoding is based on phase gradients of the transmit radio frequency (RF) magnetic field (B1). The TRASE technique is useful for building low-field and low-weight MRI systems, and is therefore suitable for space MRI applications.
The objective of the three studies described in this thesis was to further advance the TRASE technique, using both simulation and experimental approaches in 1-dimensional (1D), 1.5D, and 2D TRASE, for applications in space MRI prototypes. Prior to this work, TRASE had been successfully used to obtain clinically relevant 2D images of the wrist, using a Helmholtz-Maxwell coil array design. However, the image resolution was restricted due to coil efficiency and complexity, and further work was underway to develop a new RF coil design for TRASE. A Twisted Solenoid (TS) coil design for TRASE applications was being developed by the team at the University of Alberta (UofA). Even so, many aspects of the TRASE MRI technique remained unstudied but were crucial to the development of portable MRIs using TRASE.
The first of the three studies presented here focused on a (simulation-based) 1D TRASE pulse sequence optimization scheme. This study investigated the performance of a set of variants of a 1D TRASE sequence under conditions of |B1| errors. Results showed that, using optimum transmit pulse phases, high quality image encoding is achievable over ~90% of the Nyquist field-of-view (FOV) for a practically realizable variation in B1 amplitude of 11%. This improved significantly upon the performance of a previously-reported sequence which generated ~75% usable FOV within the Nyquist FOV. The results from this work can also be expanded for use in 2D TRASE imaging since a 2D TRASE pulse sequence uses a series of 1D TRASE echo trains.
Prior to the second study presented here, the first MRI images from a prototype Space MRI in our lab, the "Owl" MRI, had already been obtained using 1D TRASE encoding in combination with natural slice selection. For this, 1D TRASE encoding was performed at multiple B0 frequencies (to achieve natural slice selection based on the in-built B0 gradient), and the subsequent 1D TRASE images were stacked to generate a 2D image (referred to as 1.5D TRASE imaging). With the newly developed TS coil design, work on the Owl MRI presented in this second study here was based on two objectives: 1) Obtain 1.5D TRASE images from the Owl MRI using a new transmit coil array composed of a TS coil and a Saddle coil pair (replacing the original Cube coil set), and 2) Use the Owl MRI as a developmental prototype for the Merlin MRI (another Space MRI prototype that was being developed for zero-G flight tests). In this second study, one of the first 1.5D TRASE images from the Owl MRI using the second generation Owl transmit coils is presented.
Motivating the third study, a two-coil TS transmit array was being implemented for 1D TRASE by the UofA team. This work was followed by an attempt to perform 2D TRASE imaging using the TS coil array design. However, it became clear that imaging was negatively impacted due to inductive coupling among the concentrically placed B1 field coils. To understand the coupling issues, the third study presented here investigated the effects of coupled B1 fields on 2D TRASE imaging. Since TRASE relies on the use of multiple RF fields (B1 fields with different phase gradients) for k-space traversal, a TRASE pulse sequence requires RF pulses that are produced by switching between the transmit coils (B1 fields). However, interactions among the transmit RF coils can cause un-driven coils to produce unwanted B1 fields that impair the spatial encoding. The purpose of the third study was to investigate the effects of B1 field coupling using Bloch equation-based simulations and to determine the acceptable level of B1 field interactions for 2D TRASE imaging. The simulations show that 2D TRASE MRI (using a three-coil setup) displays ideal performance for pairwise coupling constant lower than k = 0.01 while having acceptable performance up to k = 0.1. This translates into S12 decoupling requirements in the range of ~(-50 dB to -30 dB) required for successful 2D TRASE MRI in this study. This result is of crucial importance for designers of practical TRASE transmit array systems.
The studies described in this thesis were all part of the the work undertaken at the Space MRI Lab, Usask, to develop MRI technology for space use. A novel research initiative such as building MRIs for space is usually a dynamic endeavour. As progress is made, new challenges are discovered which steer the path towards the final objective. The research work presented in this thesis is not different. It represents an array of studies that were conducive to the overall process of developing the technology for space MRIs using TRASE.
Description
Keywords
TRASE MRI, Space MRIs
Citation
Degree
Doctor of Philosophy (Ph.D.)
Department
Biomedical Engineering
Program
Biomedical Engineering