Solving transient problems in ultra-low-field MRI

Abstract

Magnetic resonance imaging (MRI) is widely used in clinical applications as well as in research. While the state of the art of MRI has developed towards using multiple-tesla magnetic fields, another approach has emerged, where the signal is measured in a magnetic field on the order of Earth's magnetic field (~ 100 µT). Such ultra-low-field (ULF) MRI is made possible by highly-sensitive magnetic-field sensor technology based on superconducting quantum interference devices (SQUIDs). For increasing the signal-to-noise ratio (SNR), the sample is pre-polarized in a larger field (~ 100 mT) before each acquisition cycle. ULF MRI has interesting characteristics including unique contrast mechanisms, safety, silent operation, and potential in low cost and in compatibility with other electromagnetically sensitive technology. MRI acquisition is based on measuring the nuclear magnetic resonance (NMR) resulting in the sample from applying various magnetic field pulses. Pulsing magnetic fields, especially the pre-polarizing field, however, causes a number of problems, mainly related to unwanted electromagnetic induction to other parts of the measurement apparatus and its surroundings. Eddy currents induced in the walls of the magnetically shielded room (MSR), which is required for low-noise measurements, are a particularly serious issue. The eddy currents generate a transient magnetic field, disturbing data acquisition and affecting the NMR. In this Thesis, I discuss the problems of pulse-induced transient effects and what they have in common. I adapt theory of linear systems to the effects and present methods and approaches for solving them. In particular, I derive a quantitative theory of eddy currents and magnetic shielding by them in the walls of an MSR. Some of these methods and approaches were applied to the ULF-MRI system at University of California, Berkeley. A number of upgrades were made to the system to solve several serious transient issues that made imaging impossible with a new coil setup meant to significantly improve the image quality. Among these upgrades, a new MSR was designed and constructed to reduce transient eddy currents. Time-dependent measurements of the eddy-current patterns in the MSR walls show excellent agreement with calculations based on the presented theory. Imaging tests show that the upgrades were successful. For transient suppression, I also present a method in which additional time-varying pulses are applied in the system to spatiotemporally couple to the transient effects, providing highly flexible means of simultaneously canceling transients at different time scales. A simulation study is presented, suggesting that this method can provide much better results than any pre-existing method.