Although diffusion MRI enables to probe the detailed structure of brain tissue, it is still challenging to investigate small-scale structures, such as brain cortex (CTX). Very recently, it has been shown that a navigated point-spread function (PSF)-based approach¹ at ultra-high field (≥7 T) allows measuring the diffusion properties even in brain cortex. Based on this approach, we further investigate cortical depth dependence of the diffusion properties with very high (0.8mm isotropic) resolution and overanalyse the results for the whole human brain in-vivo.

Data acquisition. A male volunteer was scanned at 7T (Siemens Healthcare, Germany). Navigated point-spread function (PSF) scans¹ without (b-value = 0) and with 12 different diffusion-weighting (DW) gradients (b-value = 800 s/mm²) were acquired with Stejskal-Tanner diffusion encoding using a 32-channel head coil (Nova Medical, Wilmington, USA). Imaging protocol parameters were: TR/TE=3560/49 ms, 41 slices (covering 3.28cm), slice thickness=0.8 mm, field of view (FOV)=192² mm², matrix size=240² (=0.8mm³, isotropic resolution). Each PSF dataset was acquired with an acceleration factor of 3 in the PSF-PE dimension (corresponding to 80 repetitions or averages) and with a reduced resolution factor of 4 in the EPI-PE dimension (resulting in a matrix size of 60 in the EPI-PE coordinate). No cardiac gating was used and the total scan time for all PSF scans was about 55 minutes. This experiment was repeated four times to cover the entire brain volume. Anatomical imaging included MP-RAGE and 3D GE acquisitions with an isotropic resolution of 0.6 mm.

Image processing. After bias field correction of MP-RAGE², brain segmentation (gray matter cortex (CTX), white matter (WM) and cerebrospinal fluid (CSF)) was performed³. To generate a whole brain volume, four different volumes from separated scans covering different regions of the brain were co-registered to the MP-RAGE image. A series of lamellae from the WM-CTX to the CTX-CSF interface was created by the surface expansion method³ in order to investigate the cortical depth dependence of diffusion properties including fractional anisotropy (FA) and mean diffusivity (MD) (Fig.1). After diffusion tensor imaging (DTI) fitting⁴, the DTI measures within each lamella were averaged over the segments of the Desikan-Killiany (DK) atlas⁵. Marginal distribution of DTI invariants were mapped onto inflated CTX of the brain and visualized for every lamella level.

A significant drop in FA value is clearly seen near the WM-CTX and CTX-CSF interfaces (Fig.2a) relative to the FA in WM. In contrast to the MD in WM, a rise of MD is observed at the same locations (Fig.2b). The lamella patterns of DTI measures seem to be very similar over all DK segments of the whole CTX. The spatial specificity of FA (Fig.3a) and MD (Fig.3b) for every lamella is outlined either by sulci or gyri if only marginal distributions are visualized.

1. The spatial FA distribution may be caused by more hindrances depending on the CTX folding leading to higher FA in sulci. Correspondingly, higher MD may represent reduced density of CTX tissue in gyri.

2. We hypothesize that effects of fiber bending can lead to a significant drop of FA and rise of MD near the WM-CTX interface (1e, Fig2). Also it seems that the tissue at the WM-CTX interface is less dense and extracellular diffusion contributes to effective diffusion. In contrast to previous reports⁶, however, CTX depth dependence of FA, with consistently higher values in the middle CTX lamella than in the deep and superficial CTX lamella, was not observed in this study.