Efficient Simulation of Magnetic Resonance Imaging

Abstract

Simulation of Magnetic Resonance Imaging (MRI) is based on the Bloch equation. Solving the Bloch equation numerically is not difficult, but realistic imaging experiments bear a high computational burden. In his dissertation, Cristoffer Cordes presents simulation methods that exploit hardware restrictions and the common structure of MRI sequences while not enforcing any approximations. These strategies use the reoccurrence of radiofrequency pulses, partial availability of analytical solutions, a reformulation of the problem in Fourier space and finally an inclusion of the reconstruction process to perform MRI simulation in image space, titled Sequence Response Kernel approach. The algorithmic efficiencies of the methods are investigated and applied to realistic imaging experiments. The properties and potential of the algorithms are exemplified, with an emphasis on the Sequence Response Kernel approach. This book is aimed at physicists and mathematicians with an MRI background.