Advancement of the Quantitative Blood Oxygenation Level-Dependent (qBOLD) MRI technique to improve clinical feasibility.

Abstract

Abstract Quantitative Blood Oxygenation Level Dependent (qBOLD) MRI offers a noninvasive method for measuring brain oxygenation, with potential applications in various neurological conditions. However, its clinical implementation has been hindered by methodological inconsistencies and challenges in standardisation. This thesis aims to address these obstacles and advance the clinical applicability of qBOLD techniques through three studies. First, a comprehensive scoping review of the qBOLD literature was conducted. The review revealed four main qBOLD acquisition methods: multiparametric (mpqBOLD), asymmetric spin echo (ASE), gradient echo (GRE), and gradient echo sampling of spin echo (GESSE). Notably, mp-qBOLD emerged as the most used technique, likely due to its easier implementation in clinical settings. However, significant variability in the reversible transverse relaxation rate (R2') measurements across different acquisition techniques was observed, highlighting the need for standardisation. To address this variability and enable quality assurance, the second study focused on developing and validating a new qBOLD phantom using glass microspheres. A linear relationship between R2' contrast and glass bubble volume fraction was established, and the phantom demonstrated good reproducibility in construction and MRI measurements. Crucially, R2' measurements were consistent across different qBOLD acquisitions and MRI vendors and the phantom accurately replicated known in vivo R2' values for human brain tissue. However, challenges arose in matching the irreversible transverse relaxation rate (R2) values to the human brain range.

The third study explored a combined hyperoxia-BOLD and mp-qBOLD (hmpqBOLD) approach to improve oxygen extraction fraction (OEF) estimation in a clinically translatable manner. This combined method overestimated OEF compared to an established technique, with values exceeding normal physiological ranges. In conclusion, this thesis has made important contributions towards addressing key challenges to the clinical implementation of qBOLD imaging, laying a solid foundation for future advancements in quantitative oxygenation imaging and its translation to clinical practice.