@phdthesis{item_3359117,
  title = {{Optimization of radiofrequency coils for human brain magnetic resonance spectroscopy at ultra-high field strength}},
  abstract = {Magnetic resonance spectroscopy (MRS) is a non-invasive and non-ionizing technique to acquire localized spectra of metabolites in vivo. With increasing static magnetic field strength B0, the spectral separation of the metabolites and the signal-to-noise ratio (SNR) of the spectrum increase. Consequently, the number of detectable metabolites and the spatial specificity are enhanced at ultra-high fields (B0 $\geq$7 T). At the same time, the wavelength of the radiofrequency (RF) field is decreased. For proton spectroscopy at ultra-high fields, the wavelength of the RF field in tissue is smaller than the typical dimension of a human head. From the perspective of electromagnetic theory, this means that a quasistatic approximation of Maxwell\textquoterights equations is not valid anymore and the electromagnetic field must be calculated with the full system of coupled partial differ- ential equations. Therefore, RF coil designs based on the quasistatic approximation, such as the birdcage coil or loop-only receive arrays, have suboptimal performance at ultra-high fields. This PhD project explored the optimization of RF coils for ultra-high field MRS. The optimization was based on an equivalent surface current distribution surrounding a human head model. It could be shown, that the equivalent surface current distribution can be separated into curl- and divergence-free components. The full-wave electromag- netic field problem was solved by a newly developed dyadic Green\textquoterights functions approach. As a first optimization goal, the SNR was maximized in a spherical- and later in a realistic human head model. By optimizing the complete set of curl- and divergence-free surface current components, an upper threshold for the achievable SNR of any receive array could be calculated; this so-called ultimate intrinsic SNR (UISNR) was studied at all practically relevant B0 field strengths regarding human head applications. The UISNR increased superlinearly with B0 in central regions of the human brain. In a next step, the SNR optimization was done separately for curl- and divergence-free current components. This yielded a direct performance measure of how close loop-only and dipole-only receive arrays were able to approach the UISNR in the human head. Based upon this analysis, field strength specific design guidelines for RF receive arrays were deduced. In conclusion, at ultra-high field strength a combination of loop and dipole elements is necessary to achieve the best possible SNR at any position in the human head. As a second optimization goal, the coupling of multi-channel RF arrays was mini- mized. For that, a fast analytical model describing the complex mutual coupling between two surface loops was introduced. To understand and eliminate both electric and mag- netic coupling between the loops, the influence of the loop geometry and loading by the sample was systematically examined. For the first time, it was demonstrated that at 400 MHz it is possible to eliminate both, electric and magnetic coupling simultaneously by proper adjustment of the loop width and overlap. A fully decoupled two channel prototype array was constructed having superior transmit and receive performance over a previously used gapped design.},
  school = {Eberhard-Karls-Universit{\"a}t T{\"u}bingen},
  address = {Tübingen, Germany},
  year = {2017},
  slug = {item_3359117},
  author = {Pfrommer, AM}
}