Advancing Ultra-High Field MRI functional and structural applications
Magnetic resonance imaging (MRI) has shown to be a valuable tool for studying the human brain, allowing in-vivo visualization of structures and anatomy in great detail, especially at Ultra-High field strengths (â‰¥ 7T). MRI is not limited by anatomical and structural information. It can study the brainâ€™s anatomy, functionality, connectivity (functional and structural), and chemical metabolism. Functional MRI (fMRI), for instance, enables the investigation of brain function mechanisms in-vivo with a non-invasive advantage compared to other tools. The present thesis focuses on advanced MRI techniques for ultra-high field strength (â‰¥ 7T), specifically for neuroscience applications. Combined with the higher field strength, these techniques provide better imaging quality and precise brain activity measurement. For example, high-quality anatomical T1 weighted images are essential for several MRI applications, notably, to serve as an anatomical reference in fMRI and gray matter segmentation. Unfortunately, increased field strength also induces non-uniformities in the transmit field (B1+) that can compromise image contrast non-uniformly. One of the goals of the present thesis was to investigate new strategies to overcome this issue. Regarding the functional brain investigation, the gradient-echo (GRE) is the typical method of choice for fMRI applications. Despite its high sensitivity to deoxyhemoglobin variations and widespread availability, the gradient-echo (GRE) BOLD signal is predominantly driven by the large draining vessels resulting in a limited spatial specificity, especially for 7T or higher field strength applications in which the BOLD sensitivity (susceptibility effect) is higher compared to lower static field scanners. In this context, we investigated an alternative fMRI method called vascular space occupancy (VASO) that promises higher spatial specificity than the typical GRE BOLD. To achieve the aim of this thesis, we used four approaches (chapters 2-5). We first evaluated the universal pulses capabilities in combination with a more adiabatic inversion pulse (TR-FOCI) to mitigate the UHF inhomogeneity problem, evaluating the contrast enhancement of the T1-weighted image from the MPRAGE sequence in comparison to the T1-weighted image of the self-bias MP2RAGE sequence (Chapter 2), and we found that these combinations can mitigate the inhomogeneity problem and increase the image quality of T1-weighted MPRAGE. For the second part of this thesis, we focused on the VASO fMRI approach; for our first fMRI assessment (Chapter 3), we implemented the Slab Selective Slab Inversion (SS-SI) VASO sequence. We investigated the linearity behavior of the VASO-CBV responses, comparing their behavior with respect to the movement rate in the motor cortex with BOLD responses. We observed a strong linear relationship between VASO-CBV and BOLD responses. We used response selectivity measurements to investigate the cortical activation with sub-millimeter VASO-CBV and BOLD fMRI data during individual finger movement (Chapter 4). We found that the higher vascular specificity of VASO-CBV fMRI results in higher response selectivity or less vascular overlap than BOLD imaging. Finally, we investigated how VASO-CBV compares to BOLD fMRI for cognitive neuroscience applications in the visual cortex (Chapter 5). We show similar eccentricity and polar angle maps. Likewise, the pRF size estimates were similar between VASO-CBV and BOLD. This work brings the ability to improve image quality and overcome typical challenges of ultra-high field MRI for both anatomical and functional applications. I hope that the neuroscience and technical fundaments discussed here can further contribute to the developing field of MRI and neuroimaging.