Investigation on the anatomical and physiological bases of functional magnetic resonance imaging signals at high magnetic fields

Functional magnetic resonance imaging (fMRI) using blood oxygenation level dependent (BOLD) contrast is one of the most powerful and widely used tools to image functional activities of the brain in vivo. The emergence of the fMRI technique has revolutionized cognitive neuroscience with the combination of noninvasiveness and high spatial and temporal resolution of fMRI. However, the mechanism and signal source of the BOLD fMRI contrast is not fully understood. The goal of this thesis is to investigate the anatomical and physiological bases of fMRI signals at high magnetic fields.; In this thesis, the contribution of intravascular signals from large veins to BOLD contrast was examined to investigate anatomical basis for the fMRI signals using a diffusion-weighting (flow-crushing) method based on the fact that the diffusion-weighting selectively attenuates signals from fast-moving vascular spins. Spin-echo (SE) technique was employed to eliminate the extravascular component of BOLD signals from large veins. The large vascular contribution to SE BOLD fMRI was negligible at 9.4 T, which is consistent with the very short T2 of venous blood compared to spin echo time.; To investigate physiological basis of fMRI signals and to characterize the BOLD fMRI signals at high magnetic field, cerebral blood flow (CBF) fMRI and SE BOLD fMRI were measured simultaneously and compared in terms of statistical significance level of activation and relative signal changes. The statistical significant level and relative signal changes of SE BOLD fMRI were linearly correlated with those of CBF fMRI to the level of individual pixel. The quantitative relationship between relative SE BOLD and CB17 signal changes suggests that SE BOLD fMRI can provide quantitative index of functional activity.; Finally, the relationships among arterial and venous cerebral blood volume (CBV) changes and CBF change during hypercapnia were investigated to elucidate the mechanism underlying the BOLD contrast. The relationship between relative venous CBV and relative CB17 was linear and the slope was 0.15, which accounts for 36% of total CBV change. The contribution of venous CBV change to BOLD contrast was about 50% of the expected contribution that was estimated based on total CBV changes.