Pushing human neuroscience past 7T: cerebellar imaging at 9.4T

Motivation: High SNR at B0>7T could aid in imaging small structures like those found in the human cerebellum, but neuroscientific feasibility is unclear. Goal(s): To visualize cerebellar structure and function in-vivo with a 9.4T protocol suitable for neuroscientific experiments. Approach: 6 individuals were scanned with a cerebellum-tailored MP2RAGE (resolution=0.4mm isotropic) and BOLD-weighted 3D-EPI runs (resolution=0.8/1.0mm isotropic) and a generalized B1+-shim. Results: The scan sessions (1hour) were feasible for neuroscientific experiments. Sufficient contrast for gray-white matter segmentation, and consistent fMRI responses were found for all participants. Impact: Reliably visualizing the cerebellum requires pushing past currently-feasible resolutions with B0=7T or lower. We demonstrate that neuroscientific experiments can be performed in the human cerebellum at 9.4T benefitting from elevated SNR and BOLD sensitivity.