Daniel Hänelt | Cortical depth-dependent fMRI in the human visual cortex at ultra-high magnetic fields

Advances in functional magnetic resonance imaging (fMRI) promise detailed non-invasive probing of neural activity in the human brain. However, since the measured MR signal depends on the hemodynamic response following neural activation, we have only indirect access to the signal source of interest. Especially at higher resolutions, the actual source of the measured signal gets important which depends on the vascular architecture of the cerebral cortex. The venous side of the cortical vascular tree consists of a capillary bed within the cortex and larger intracortical veins draining the blood to large pial veins located on top of the cortical surface. To be closest to the source of neural activation, it is desirable that the acquired MR signal is most sensitive to the capillary bed. Different fMRI methods like GE-BOLD, SE-BOLD or VASO inherently differ in terms of their sensitivity and specificity to the wanted signal source. In our work, we try to get further insights about the signal dependencies of different fMRI methods at ultra-high magnetic fields (≥ 7 T) to the macrovasculature, especially to large pial veins. Currently, we run two experiments which both target the human visual cortex as model system. In one experiment, we map ocular dominance columns (ODCs) in the primary visual cortex (V1) and aim to infer the vascular induced spread of the MR signal by analyzing the topography of ODCs across cortical depth. In another experiment, we use the retinotopical ordering of V1 to get a more general understanding of the signal spread across cortical depth.