Scientist Center for Neuroscience and Regenerative Medicine, HJF
Abstract Text: INTRODUCTION Imaging brain tissues properties at the mesoscopic scale is an invaluable method for studying normal and abnormal brain function particularly traumatic brain injury (TBI) which has been so far invisible to standard MRI scans. However, the low sensitivity of NMR makes it challenging to achieve native mesoscopic voxel resolution. Diffusion tensor distribution (DTD) MRI models the voxel as an ensemble of microscopic diffusion tensors thereby enabling the measurement of diffusion processes at the mesoscopic and microscopic scales, orders of magnitude smaller than the nominal voxel size. In this study, we used state-of-the-art 300 mT/m gradients with high imaging acceleration factors to acquire multiple diffusion encoded (MDE) data of a whole human brain in vivo at 1.2 mm isotropic resolution and estimate DTDs. The high spatial resolution of the scans reduces the structural complexity of the tissue within each voxel and allows us to probe intracortical heterogeneity effectively.
METHODS The diffusion weighted signals were acquired using a novel double PFG pulse sequence called interfused-PFG, a.k.a. iPFG, with EPI readout and multi-band (MB) slice excitation, which uses efficient trapezoidal gradients with well-defined diffusion times. The sequence is also immune to concomitant gradient field artifacts known to bias DTD results. Diffusion encoding was obtained using rank-1 and rank-2 b-tensors. The two independent gradients in the diffusion block were estimated numerically to yield a set of 216 diffusion encoding b-tensors with uniform distribution of sizes (b = 0 – 2,500 s/mm2), shapes, and orientations for an unbiased estimation of the DTD.
The DTD was assumed to be derived from a normal tensor variate distribution (NTVD) whose samples are constrained to be positive definite (CNTVD) to ensure physicality, and estimated using Monte Carlo method. CNTVD requires only rank-1 and rank-2 b-tensors for DTD estimation and yields a compact, maximum entropy distribution. The estimated DTD was used to synthesize micro-diffusion tensors within each voxel whose size, shape and orientation heterogeneity were quantified using measures derived from their eigenvalue and eigenvector distributions. The ODFs were computed for the mean tensor (i.e., DTI-ODF) and the distribution of tensors (i.e., DTD-ODF) to perform streamline tractography in MRTrix software environment.
MRI data in a healthy volunteer was acquired on a 3T scanner (MAGNETOM Connectome, Siemens Healthineers) with 300 mT/m peak gradient strength and a 200 T/m/s slew rate using 64-channel coil. Whole-brain DTD MRI data was acquired using the following parameters: δ\τ_m\Δ = 9\9\33 ms, FOV=180 x 180 x 148 mm, MB factor = 3, GRAPPA acceleration factor = 2, 3 averages, TR\TE = 5,500\82 ms, and a 1.2 mm isotropic spatial resolution.
RESULTS AND DISCUSSION The results showed variability in the mesoscopic heterogeneity across different areas of the cortex especially in the visual cortex (Brodmann area, BA 17) and the operculum (Brodmann area, BA 44). In BA17, there was high orientation heterogeneity but low size heterogeneity which may be due to powder averaging of equisized cells and/or splaying of white matter fibers as they enter the cortex. In BA 44, we observed large values of size, shape and orientation heterogeneities with the size and shape skewness measures displaying intracortical contrast akin to layers. This tripartite layering structure aligns with the known cytoarchitectonic features in this area. The DTD contrast was also observed in sub-cortical regions and various white matter regions. The orientational heterogeneity of white matter was also captured in DTD tractography result which showed complex fiber organization in the DTD derived tracts absent in the DTI-ODF derived tracts.
CONCLUSION We have introduced a suite of new diffusion MRI tools to study mesoscopic heterogeneity in whole brain using DTD MRI. While it is now possible to identify layered structures in the cortex, a longer-term goal of this DTD project is to be able to image the gray/white matter interface and other interfacial areas reliably and clearly, like periventricular areas, that are known to be implicated in blast and other forms of TBI.
