Image Encoding and Reconstruction for Portable Magnetic Resonance Imaging

Magnetic Resonance Imaging (MRI) is a successful imaging tool, but due to high cost, high weight and complexity of equipment, MRI is not currently as easily accessible clinically as desired. By making MRI cheaper, lighter, less complex and therefore potentially portable, it can become more widely accessible. A portable MRI system can be used in primary health care, operating and emergency rooms, car and air ambulances, sport and war facilities and remote regions (including outer space). The objective of this study was to show the feasibility of reconstructing MRI signals generated by two different portable MRI systems. The first portable MRI, known as radiofrequency (RF) phase encoded MRI, encodes spatial information through the use of a non-linear spatially varying RF transmit ($B_1$) phase. The second portable MRI, known as rotating field MRI, encodes information through non-uniform radially varying main magnet field ($B_0$). The fact that there is no need for gradient coils in both systems, leads to a smaller, lighter and more affordable MRI than most conventional systems. In RF phase encoded MRI, since the $B_1$ phase spatially varies non-linearly, using Fourier transform (FT) to reconstruct images results in distorted images. Therefore, regularized least squares inversion was used in place of the usual FT. The RF phase encoding coil generates an inhomogeneous $B_1$ field that leads to RF pulse imperfection in terms of flip angles produced versus flip angles intended. Composite pulses were therefore used to minimize the effect of RF transmit field inhomogeneity on tip angles. In rotating field MRI, a Halbach magnet was used to generate a non-uniform radially varying $B_0$ field to encode information in the radial direction. For encoding information in the angular direction two separate Saddle RF receiver coils were used. The main magnet and receiver coils are fixed relative to each other, but both rotate around the object. A regularized least squares inversion (LS) method followed by total variation (TV) techniques were used to reconstruct the images. MRI simulation signals encoded in RF transmit field with non-linearly varying spatial phase may be accurately reconstructed using regularized LS method thus pointing the way to the use of simple RF coil designs for RF encoded MRI. Also, my results from simulation and experimental data, indicated the feasibility of reconstructing images from rotating field MRI. I have made progress in the realization of a novel approach to different MRI systems that do not rely on active magnetic gradient fields. These two methods can be combined to encode information in 3 dimensions (3D) in the future, for example inhomogeneous $B_0$ field can be used for slice selection and RF phase encoding can be used to encode information in the plane