NEURORADIOLOGYT he human cerebellum has approximately one-eighth the volume of the neocortex but a cortical sheet with a similar surface area (1). The cerebellar cortex is a prevalent lesion site for various disease processes, including multiple sclerosis (2,3), spinocerebellar ataxias (4), and alcoholism (5).The cerebellar cortex has a regular circuit characterized by the granular cells (granular layer: 250-350 μm wide) that output to the Purkinje cells (Purkinje layer: 12 μm) through the parallel fibers (molecular layer: 250-350 μm). Beams of parallel fibers underlie the characteristic cerebellar mediolateral folds (500-1000 μm) (6). Despite its importance, this mesoscopic scale of the cerebellar architecture is underexamined in humans: Current in vivo techniques are limited to gross anatomic features, such as the main cortical branches, due to inadequate resolution (1). As a result, clinical measures successfully used in the cerebral cortex, such as cortical thickness, volume, and myeloarchitecture/function correlates, have not been applied with high fidelity in the cerebellum (1). This hinders examination of the cerebellum in disease.Herein, we combined methodologic advances in image acquisition and analysis to image the cerebellar cortex, including its foliations and lamination, with use of in vivo 7.0-T MRI to improve the signal-to-noise ratio and signal-to-noise efficiency. We modified two MRI pulse sequences targeted at imaging the cortical surface and intracortical lamination, respectively, to translate this higher signal-to-noise ratio to spatial resolution Background: The human cerebellum has a large, highly folded cortical sheet. Its visualization is important for various disorders, including multiple sclerosis and spinocerebellar ataxias. The derivation of the cerebellar cortical surface in vivo is impeded by its high foliation.Purpose: To image the cerebellar cortex, including its foliations and lamination, in less than 20 minutes, reconstruct the cerebellocortical surface, and extract cortical measures with use of motion-corrected, high-spatial-resolution 7.0-T MRI.
Materials and Methods:In this prospective study, conducted between February 2021 and July 2022, healthy participants underwent an examination with either a 0.19 × 0.19 × 0.5-mm 3 , motion-corrected fast low-angle shot (FLASH) sequence (14.5 minutes) or a whole-cerebellum 0.4 × 0.4 × 0.4-mm 3 , motion-corrected magnetization-prepared 2 rapid gradient-echo (MP2RAGE) sequence (18.5 minutes) at 7.0 T. Four participants underwent an additional FLASH sequence without motion correction. FLASH and MP2RAGE sequences were used to visualize the cerebellar cortical layers, derive cerebellar gray and white matter segmentations, and examine their fidelity. Quantitative measures were compared using repeated-measures analyses of variance or paired t tests.Results: Nine participants (median age, 36 years [IQR, 25-42 years; range, 21-62 years]; five women) underwent examination with the FLASH sequence. Nine participants (median age, 37 years [IQR,[34][35...