King's College London

Research portal

Optimized Variable Flip Angle Methods for Single Pool MRI Relaxometry

Student thesis: Doctoral ThesisDoctor of Philosophy

Magnetic Resonance Imaging (MRI) is routinely used as a highly, soft-tissue sen-sitive qualitative modality. Thus, although widely used as a first line of investiga-tion for both radiological diagnosis and treatment monitoring of neurological dis-ease, almost all assessments are based on images presented in arbitrary units. With this in mind, there is a growing interest in quantitative MRI as a potential route to less subjective diagnosis and to allow cross site comparison studies. Key MR parameters are the proton density M0 and relaxation times T1 and T2 which are strongly associated with tissue integrity. This absolute tissue specific measurements, are expected to overcome inter-site bias in multi-centre studies as opposed to conventional M0, T1 and T2 weighted images whose use is still controversial.
Unfortunately, gold standard methods for estimating relaxation times are two dimensional acquisitions based on spin-echo processes which require long acqui-sition times. On the contrary, many gradient echo techniques, such as Variable Flip Angle (VFA), Driven Equilibrium Single Pulse Observation of T1/2 (DESPOT), Muti-Parametric Mapping(MPM), etc, have been developed to infer tissue MR properties in clinically feasible times. However, a consensus regarding the accu-racy of each method has still to be found. One possible source for the reported discrepancies between methods, is the fact that, in biological samples, a process called Magnetization Transfer (MT) is known to influence the observed relaxom-etry measurements. To characterize tissue more fully, so called multiple-pool models have been suggested. Current clinical protocols for quantitative imaging generally fail to take MT correctly into account, and therefore produce variable results that undermine their utility as secure diagnostic methods. Quantitative MT protocols can more precisely characterise tissue, but require more data to be col-lected so are not regarded as clinically feasible.
The work here presented, built on single-compartment DESPOT relaxome-try approach and sought to increase its precision of by two means: (i) a joint system relaxometry (JSR) approach that estimates parameters in a single step using all available data; and (ii) optimizing acquisition parameters by deploying a robust design tool based on the Cr´amer-Rao lower bound (CRLB). Once this was achieved, the absolute accuracy of gradient echo methods was explored by exploring the influence of magnetization transfer effects on single-pool assump-tions. It was then hypothesised that robust relaxometry methods can be achieved by ensuring Constant Saturation of Magnetization Transfer (CSMT) effects. This was demonstrated both numerically and experimentally.
Original languageEnglish
Awarding Institution
    Award date2017


    Download statistics

    No data available

    View graph of relations

    © 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454