Abstract
Magnetic resonance imaging is a powerful noninvasive method for imaging cross-sectional slices
in the human body and brain, as well as metabolic processes therein. It capitalizes from the
magnetizability of tissue in strong static magnetic fields and the possibility to make the local
magnetization detectable with additional radiofrequency pulses, based on the physical principle
of nuclear magnetic resonance.
In 2008, magnets that produce a static field strength (
B
0
)of0
.
5or1
.
5 Tesla are the clinical
standard for routine radiological exams on humans, but this standard is apparently moving
slowly to the use of 3 Tesla magnets. However, there is a drive to even higher field strengths
in order to explore its advantages for clinical and research imaging. In 1998, the first 8 Tesla
system with an 80 cm bore was built and installed at Ohio State University, followed by the
first 7T/90cm magnet at the Center for Magnetic Research at the University of Minnesota in
1999 [28]. Several other sites installed magnets with a field strength of 7 Tesla or higher since
then, and in 2008, three operational 9
.
4 T systems for human research exist, one of them at the
Magnetic Resonance Center of the Max Planck Institute for Biological Cybernetics in T
̈
ubingen,
Germany. The main motivation for high-field magnets in MRI is the theoretical proportionality
of the signal-to-noise ratio (SNR) to the
B
0
field magnitude, because the increased signal from
the sample allows for scanning with a higher spatial or temporal resolution. SNR-demanding
techniques, as for example fMRI, parallel imaging methods or imaging of low
γ
nuclei strongly
benefit from the increase in field strength. In addition, MR spectroscopy always aims at higher
field strengths because the chemical shift dispersion increases with
B
0
.
However, some serious challenges exist with human high-field imaging. The most dominant of
them are related to the increasing frequency of the magnetic field
B
1
that must be generated
by the RF coils since the resonance frequency grows proportional to the static magnetic field
B
0
. The decreased wavelength and penetration depth of the radiofrequency field complicate the
creation of a homogeneous circularly polarized magnetic RF field inside the sample, leading to
inhomogeneous flip angle maps and therefore inhomogeneous images. In addition, the power
deposition or specific absorption rate (SAR) due to the accompanying electric field
E
1
increases
at high frequencies, and this can possibly lead to a dangerous rise in temperature in the imaged
subject. Unfortunately, no direct methods exist to map the local RF electric field
in vivo
,and
therefore a direct measurement of SAR is not possible.
Numerical simulations have proved to be useful in addressing these issues. In the last couple
of years, the finite-difference time-domain (FDTD) method was used by a few researchers to
accurately calculate electromagnetic fields produced by RF coils in the presence of complex
shaped biological tissue. Their work greatly improved the understanding of field behavior in
high-field MRI and lots of universally valid results could be extracted from their efforts.
This work mainly focuses on the precise simulation of electromagnetic fields produced by phased
array coils for human brain imaging at 9
.
4 Tesla. Chapters 1, 2 and 3 deal with the basics
of NMR, coil design and the numerical algorithm used, while the materials and methods are
explained in chapter 4. The results of the simulations for RF coils at 9
.
4 Tesla are presented in
chapters 5 and 6, the results of the simulations for coils at 16
.
4 Tesla can be found in chapter 7.
Chapter 8 finally gives a brief discussion about the findings of this work.
