Neuroimaging at ultra-high spatiotemporal resolutions: line-scanning fMRI

Abstract

To be able to investigate the information processing across cortical depth non-invasively, fMRI techniques need to be improved to allow, at the same time, high spatial and temporal resolutions. In this thesis, line-scanning fMRI has been extensively studied as an extreme fMRI approach for ultra-high spatiotemporal resolution. Its one-dimensional nature requires new acquisition strategies as well as customized approaches for data reconstruction and analysis. The reader will be guided through different steps involved in development of line-scanning fMRI technique and its applications, after a short introduction on current fMRI acquisition methods for functional connectomics. Chapter 2 contains a literature review on the current acquisition techniques for resting state fMRI, from the most common approaches for resting state acquisition strategies, to more recent investigations with dedicated hardware and ultra-high fields. This chapter offers an opposite point of view compared to the other chapters, going from an extended, whole brain FOV acquisitions suitable for connectivity analysis to the extremely reduced FOV of the lines, analysed with a strong focus on task-based experiments. In Chapter 3, the first implementation of gradient-echo line-scanning is presented: we analysed the quality of line-scanning data acquisition and optimized the reconstruction strategy. We applied the line-scanning method in the occipital lobe during a visual stimulation task, showing BOLD responses along cortical depth, every 250 mm with a 200 ms repetition time. As proof-of-concept, we compared t-statistical values from line-scanning with 2D gradient-echo echo planar imaging BOLD fMRI data with the same temporal resolution and voxel volume, and showed a good correspondence between the two. In Chapter 4, a comprehensive update to human line-scanning fMRI is introduced. First, we investigated multi-echo line-scanning with different protocols varying the number of echoes and readout bandwidth while keeping the TR constant. We also implemented an adaptation of NOise reduction with DIstribution Corrected principal component analysis (NORDIC) thermal noise removal for line-scanning fMRI data. Finally, we tested image-based navigators for prospective motion correction and examined different ways of performing fMRI analysis on the timecourses, which were influenced by the insertion of the navigators themselves. Together those changes improve the robustness of the line-scanning method. In Chapter 5, an alternative contrast for line-scanning is proposed. Spin-echo is a natural candidate for line-scanning, due to its innate properties of sharp line selection and the microvascular selective functional contrast. However, we could not detect any activation in the visual cortex after a very strong visual task, hence we concluded that further improvements in the spin-echo line-scanning acquisition need to be introduced before it can be applied for neuroscientific purposes. In Chapter 6, an example of line-scanning application is presented: we explored HRF changes in visual cortex across cortical depth in two age groups. We also bridge ageing HRF changes obtained through line-scanning with more conventional fMRI acquisitions at high-field. The thesis concludes with a general discussion in Chapter 7.