The objective of the proposed research is to map the organization of human white matter (WM) with cutting-edge ex vivo imaging technologies. This work will produce microscopic-level information on several long-range WM projections, as well as a more targeted mapping of circuits that serve the prefrontal cortex (PFC). These circuits are of particular interest in psychiatric applications but they have been heretofore mapped extensively only in non-human primates. Specifically, recent work with tracer injection studies in macaque monkeys has established how small fiber bundles that originate in different areas of the PFC reach their destinations by using the large WM pathways, such as the cingulum bundle (CB), corpus callosum (CC) and uncinate fasciculus (UF), as their conduits. For example, the UF is composed of several sub-bundles: some that follow the entire trajectory of the UF and others that join the UF only for part of its trajectory to then jump off and join other large pathways, like the CB, CC, etc. Mapping these distinct components of larger WM pathways in the human brain is challenging both in vivo and ex vivo. Invasive injection studies are not applicable to humans and conventional 2D histological techniques like myelin staining cannot be used to infer the 3D orientation of axon bundles. Diffusion MRI (dMRI) can provide estimates of these orientations indirectly, by measuring the diffusion of water molecules through the WM. However, it is prone to errors in areas of complex WM architecture and requires validation by an independent source of measurements. In this work we will combine high-resolution, high-SNR ex vivo dMRI with polarization-sensitive optical coherence tomography (PS-OCT) in post mortem human brains to extract microscopic information on WM areas that confound conventional dMRI, and to perform a detailed mapping of the projections of the PFC. We will take advantage of the MGH Connectom scanner, a unique instrument that can achieve 8 times stronger diffusion-encoding gradients than routine scanners, and was designed specifically for high-SNR, high-resolution dMRI. We will develop a specialized receive coil array for imaging ex vivo human brains on the Connectom scanner, which will allow us to collect whole-brain dMRI data with unprecedented resolution and SNR. The gold-standard dMRI and PS-OCT data produced by this work will be used to construct a novel atlas of WM anatomy, which will be incorporated in a tool for automated global probabilistic tractography, building on prior work by the PI. This tool will allow both the new, detailed taxonomy of smaller WM bundles, as well as the classical definitions of large WM pathways, to be reconstructed automatically from routine-quality in vivo dMRI data that can be collected on conventional scanners. The proposed work promises to advance our understanding of the organization of human neurocircuitry; to move human dMRI studies from the current view of a WM pathway as a single bundle to one where multiple smaller bundles merge on and off a pathway at different parts along its trajectory; and to provide the tools for studying this detailed WM taxonomy using routine neuroimaging data.
This project will combine cutting-edge diffusion MRI and optical imaging technologies to map complex networks of the human brain that have been implicated in psychiatric disorders. We will use diffusion MRI and optical coherence tomography data collected at very high resolutions from post mortem human brains to build models of the pathways involved in these networks. We will incorporate these models into an algorithm for the automated and robust reconstruction of these pathways from diffusion MRI data that can be acquired in vivo, thus producing a valuable tool for researchers who use this type of data to study the structure of the human brain.
|Rosas, H D; Wilkens, P; Salat, D H et al. (2018) Complex spatial and temporally defined myelin and axonal degeneration in Huntington disease. Neuroimage Clin 20:236-242|
|Siless, Viviana; Chang, Ken; Fischl, Bruce et al. (2018) AnatomiCuts: Hierarchical clustering of tractography streamlines based on anatomical similarity. Neuroimage 166:32-45|