MRI Pulse Sequence Development and Accelerated Imaging
Abstract
Magnetic resonance imaging (MRI) of the lungs using hyperpolarized gases (e.g., 3He, 129Xe) is a powerful, non-invasive diagnostic tool for the characterization and treatment of various lung conditions, giving clinicians the ability to directly image the lung microstructure and observe gas exchange. However, non-healthy patients most likely to require pulmonary MRI scans will also present with reduced lung function, limiting their ability to hold their breaths for long periods of time (typical breath-holds are 16 sec). It is known that the optimal field strength range for pulmonary MRI using hyperpolarized gases lie much lower than clinical field strengths, between 0.1 T and 0.6 T. The low field regime offers longer T2* relaxation and higher signal-to-noise ratio for this application: these advantages facilitate certain types of imaging (e.g., longer diffusion times) and diagnostic capabilities, and the low field scanners could potentially help address a clinical need for lung imaging tools. Thus, this thesis aims to improve imaging quality of conventional proton-based MRI at low field, and accelerate pulmonary MRI.
Clinical adoption of low field MRI is hindered by the inherent low signal and resolution of 1H-based imaging; dedicated advanced imaging methods and hardware may be able to remedy these issues. The first part of this work consists of the development of the X-Centric pulse sequence for low field MRI. This sequence has an ultra-short echo time with short acquisition windows, maximizing signal for short signal-lifetime samples.
Most clinical applications of undersampled MRI are limited to 2 or 3-fold acceleration. Recently, the SIDER (SIgnal DEcay into the Reconstruction) reconstruction method has shown considerable performance for up to 10-fold acceleration, but was limited to diffusion imaging. A signal model of RF depolarization was introduced in this thesis to permit static ventilation imaging with this method, and a novel image averaging scheme was developed to significantly accelerate low field MRI whilst improving image quality: 14-fold acceleration in human lungs was made possible in this work (not yet clinically viable).
This thesis presents novel imaging techniques and tools for low field and pulmonary MRI, resulting in significant acceleration and improvements in signal and image quality.