Abstract

Magnetic Resonance Imaging (MRI) has become one of the most important medical imaging modalities over the past few decades because of its flexibility and low risk, along with other useful attributes. For traditional MRI, the static magnetic field, B_0, must be highly homogeneous. Obtaining this homogeneity can be difficult. Traditional MRI also requires linear gradient fields that are directed along the static field direction. Under these conditions a Fourier transform relationship exists between sampled data and the image to be reconstructed. In the case of an inhomogeneous static field, gradient fields that are not linear, or gradients that are not pointed along B0, there will be no Fourier transform relationship, but a linear relationship does exists and imaging is still possible. This thesis explores the possibilities of inhomo- geneous field imaging and presents the development of hardware for inhomogeneous MRI research. Two techniques for inhomogeneous imaging are derived and presented. Matrix operators are found for these two imaging methods which are analyzed using a singular value decomposition. This analysis shows that reconstructing an image from an inhomogeneous system is possible if a field map is available.

Degree

College and Department

Ira A. Fulton College of Engineering and Technology; Electrical and Computer Engineering

Rights

http://lib.byu.edu/about/copyright/