Blood Flow and Brain Function: Investigations of neurovascular coupling using BOLD fMRI at 7 tesla

Abstract

The advent of ultra high field (7 tesla) MRI systems has opened the possibility to probe biological processes of the human body in great detail. Especially for studying brain function using BOLD fMRI there is a large benefit from the increased magnetic field strength. BOLD fMRI is the working horse in neuroscience for studying brain function as it can non-invasively detect blood flow changes due to increased neuronal activity. Besides the overall increase in signal-to-noise-ratio at 7 tesla , there is an increase in the BOLD signal changes from the vasculature upon neuronal activation. In this thesis, the increased signal strength at 7 tesla allowed us to study the BOLD signal characteristics in depth, obtaining detailed spatial and temporal information of the underlying neurovascular coupling mechanisms. Also, in two studies we had the opportunity to scan patients before implantation of electrode grids (electrocorticography or intra-cranial EEG, ECoG), and thus link 7 tesla BOLD fMRI to the underlying neuronal physiology. We show that subsecond temporal differences in BOLD signal changes can be resolved as function of cortical depth, which is at millimeter spatial scale. Results also highlight the challenge that BOLD responses contain a multitude of neuronal and non-neuronal, i.e. solely blood drainage, related hemodynamics. Nonetheless, the early phase of the BOLD signal in gray matter was found to be mainly associated with microvascular hemodynamics (capillary bed), which is probably the most specific vascular marker for neuronal activity. Furthermore, by combining 7 tesla BOLD fMRI and electrophysiology (ECoG) we revealed that there is a tight neuronal correlate of the elicited hemodynamics in sensorimotor cortex. This was found to manifest at a fine spatial scale (millimeters) but also in the situation of apparent nonlinearities in both hemodynamic and neuronal responses. In conclusion, these results promise a high potential for 7 tesla BOLD fMRI to separate neuronal and non-neuronal related hemodynamic events, and hence the opportunity to elucidate brain function.