Chemical exchange saturation transfer methods for clinical magnetic resonance imaging

Thesis

Chemical exchange saturation transfer (CEST) imaging is a novel contrast mechanism in magnetic resonance imaging. CEST allows for the indirect detection of chemical groups that contain exchangeable protons. The contrast from endogenous CEST agents in the human brain is sensitive to changes in pH and protein structures. Therefore, CEST imaging offers a potential tool for the non-invasive detection of pathologies such as stroke and cancer.

However, clinical implementation of current CEST MRI sequences is hindered by a lack of contrast-to-noise ratio (CNR) per unit time. Furthermore, confounding effects which contribute to the development of CEST contrast make interpretation of this contrast difficult. This thesis examines the underlying physical effects that govern the development of CEST contrast and aims to improve clinical CEST sequences.

The ability to correctly identify CEST parameters from z-spectra was investigated through simulations of the Bloch-McConnell equations. Conventional least squares fitting methods are inherently sensitive to changes in the bulk water. Reducing the sampling of the bulk water frequencies improved the sensitivity of this method to the exchange rate.

In vitro experiments using conventional non-interleaved CEST sequences demonstrated that reduction of the longitudinal magnetisation close to the bulk water frequencies is dominated by T₁ relaxation. Minimisation of the inversions of the bulk water reduced the longitudinal component of the direct water saturation (DWS) effect. Implementation of additional acquisition events allowed for the transverse component of DWS to be used for imaging.

Interleaved CEST sequences were used in phantoms to study the effect of the interleaved readout excitations. Optimal flip angles, 𝛼, for the readout pulses fulfilled the  condition 0 < α < α^ernst and allowed compromise between optimal CEST contrast and signal-to-noise ratio (SNR). Reducing SNR in favour of CNR made the interleaved CEST sequences more prone to coherence pathway artefacts. Hexagonal spoiling in interleaved CEST reduced this problem.

Finally, in vivo experiments showed that confounding effects such as T₁ can be used as a contrast mechanism to optimise CEST sequences directly in healthy volunteers where conventional stroke contrasts are lacking.

Authors: Brand, R

• DOI: 10.5287/ora-a4a48odmg
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford