Thesis defense Sahra Fazal (Donders series 581)

11 November 2022

Promotor: prof. dr. D. Norris

The Nuts and Bolts of fMRI Acquisition protocols

In fMRI, multiple factors are involved when optimizing the acquisition protocol due to the vast parameter space. The optimum protocol is the one with a high spatial and temporal resolution within the constraints of the hemodynamic changes, whole brain coverage, and maximum sensitivity for BOLD and minimum artifacts. Some of these conditions can be achieved with the advent of online reconstruction for highly accelerated sequences and openly available MB sequences. Therefore, there is recently an inclination in fMRI toward MB. The MB/SMS protocol offers shorter TR with moderate spatial resolution. However, from an SNR efficiency perspective, there is little benefit to having TRs much shorter than T1 of the tissue of interest. Therefore, we compare a faster multiband scan with a multiple echo scan. In resting fMRI, a sensitivity gain from a short TR is gained. A denser sampling can efficiently remove physiological and structured noise from the data, thereby increasing the tSNR. The general advice would be that, in resting fMRI, the choice of protocol would be to use MB if multiple regression-based approaches are used to analyze the data. In task-fMRI, as mass univariate regression is the standard approach for analysis, MBME is the preferred protocol. We propose a novel method for fMRI data acquisition called ABC (Arterial Blood Contrast). This approach sensitizes the MRI signal to the changes in blood volume by using magnetization transfer to differentially attenuate the signal from tissue, while that from blood is largely unaffected. An increase in blood volume, such as that occurring because of brain activation, will lead to an increase in signal intensity.