Luis Manuel Colon-Perez¹, Caitlin Spindler², Shelby Goicochea³, William Triplett⁴, Mansi Parekh⁵, Eric Montie⁶, Paul Richard Carney^5,7, and Thomas Mareci^4
1Physics, University of Florida, Gainesville, Florida, United States, ²Biology, University of Florida, ³Chemistry, University of Florida, ⁴Biochemistry & Molecular Biology, University of Florida, ⁵Pediatrics, University of Florida, ⁶Science and Mathematics, University of South Carolina Beaufort, Bluffton, South Carolina, United States, ⁷Wilder Center of Excellence for Epilepsy Research, University of Florida

MRI derived measurements such as brain network measures are inherently discrete and are affected by the resolution of acquisition. Appropriate normalization techniques permit the use of DWI and tractography methods to create measurements that would allow the study of the fibrous structure within the brain. Here we present an edge weight metric used to define the connectivity strength of a network node that would allow a better understanding of local networks in the brain. With this edge weight metric, anatomical structures are defined, i.e. nodes, and the implied white matter tracts from tractography give rise to the edge.

Matteo Bastiani^1,2, Rainer Goebel¹, and Alard Roebroeck^1
1Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience, Maastricht University, Maastricht, Netherlands, ²Forschungszentrum Jülich, Jülich, Northreinwestfalen, Germany

This work focuses on reconstructing the macroscopic human cortical connectome using diffusion weighted imaging. This can be represented as a graph where cortical patches are nodes and white-matter connections are edges. The effects on estimating the most common graph measures when using single versus multi direction diffusion models, deterministic versus probabilistic tractography, and local versus global measure-of-fit of the reconstructed fiber trajectories are evaluated. We show that these choices, together with anisotropy, curvature and probabilistic threshold parameters can strongly affect connection density, small-worldness and cortical connection hubs. This is an important consideration for studies using these measures as dependent variables.

David Moreno-Dominguez¹, Alfred Anwander¹, and Thomas R Knösche^1
1Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany

Hierarchical clustering of probabilistic tractograms encodes the information of the connectivity structure at all granularity levels in a hierarchical tree or dendrogram. It might be the key to whole-brain connectivity based parcellation, where the correct number of clusters is unknown and depends on the desired granularity. The interpretation of the resulting dendrogram is not simple, due to outliers and the size of the dataset encoded, among other reasons. In this study a fast, fully hierarchical bottom-up algorithm is presented, and intelligent processing steps are introduced in order to ease the information extraction process, successfully enabling better performance of tree-partitioning algorithms.

Robert Elton Smith^1,2, Jacques-Donald Tournier^1,2, Fernando Calamante^1,2, and Alan Connelly^1,2
1Brain Research Institute, Florey Neuroscience Institutes, Heidelberg, Victoria, Australia, ²Department of Medicine, University of Melbourne, Melbourne, Victoria, Australia

When performing whole-brain fibre-tracking using Diffusion-Weighted Imaging, certain pathways in the brain can be over- or under-estimated, due to the independence of individual streamlines, and any methodological biases of the models or algorithms used. To improve the biological accuracy of connectome reconstruction, we propose a normalisation process for whole-brain fibre-tracking data, which uses the results from spherical deconvolution to selectively filter tracks from the data set, matching the track densities in high angular resolution space to the estimated intra-cellular volume fractions.

Chunlei Liu^1,2, Bing Wu¹, and Wei Li^1
1Brain Imaging and Analysis Center, Duke University, Durham, NC, United States, ²Department of Radiology, Duke University, Durham, NC, United States

A method of fiber tractography based on STI was demonstrated. STI experiments of perfusion-fixed mouse brains were conducted and magnetic susceptibility tensors were calculated for each voxel with regularization and decomposed into its eigensystem. The major eigenvector was found to be aligned with the underlying fiber orientation. Following the orientation of the major eigenvector, we were able to map distinctive fiber pathways in 3D. As a comparison, diffusion tensor imaging (DTI) and DTI fiber tractography were also conducted on the same specimens. The relationship between STI and DTI fiber tracts was explored with similarities and differences identified.

Elizabeth Zakszewski¹, Andrew S Fox¹, Jonathan Oler¹, Nagesh Adluru¹, Alexander K Converse¹, Jeffrey M Moirano¹, Ned Kalin¹, and Andrew L Alexander^1
1University of Wisconsin, Madison, WI, United States

Examinations of probabilistic tractography have identified major tracts across the brain, but little has been done to compare them to well-established tracer studies. Therefore, diffusion tensor imaging (DTI) probabilistic tractography of white matter (WM) was performed on a DTI template generated from 271 rhesus macaques. Injection sites from tracer experiments in literature were chosen as seed regions from which to begin tractography. Results demonstrated the probabilistic connectivity maps obtained on the DTI template were largely consistent with connectivity maps from the tracer studies.

Derek K Jones¹, Kat F Christiansen¹, Rosanna J Chapman¹, and John P Aggleton^1
1CUBRIC, School of Psychology, Cardiff University, Cardiff, Wales, United Kingdom

The cingulum bundle is a complex tract comprised of many different connections with trajectories of different lengths. However, this complex composition is rarely reflected in published diffusion tractography images, which most often show a continuous band of white matter that seemingly links (uninterrupted) the caudal medial temporal lobe with retrosplenial, cingulate, prefrontal, and subgenual areas. This work examined the unity of the cingulum bundle using tractography by comparing results from ROIs at different points along the tract and reveals a set of overlapping, but largely separate, connections within what is often visualised as a unified bundle.

Kawin setsompop¹, Berkin Bilgic², Julien Cohen-Adad¹, M. Dylan Tisdall¹, Boris Keil¹, Thomas Witzel¹, Yogesh Rathi³, Van J Wedeen^1,4, Elfar Adalsteinsson^2,4, and Lawrence L Wald^1,4
1Dept. of Radiology, A. A. Martinos Center for Biomedical Imaging, MGH, Charlestown, Massachusetts, United States, ²EECS, Massachusetts Institute of Technology, Cambridge, Massachusetts, ³Brigham and Women's Hospital, Boston, Massachusetts, United States, ⁴Harvard-MIT Division of Health Sciences and Technology

In this work, we demonstrate an improvement in the time efficiency of DSI acquisitions using three complementary technologies: (i) high-strength gradient coils, (ii) a Simultaneous Multi-Slice (SMS) acquisition with Blipped-CAIPI readout using a highly parallel receive coil array, and (iii) a compressed sensing reconstruction that enables undersampling of q-space. Together, these improvements allow for an acquisition of high-quality whole-brain DSI data in just over 4 minutes. While this initial demonstration focuses on DSI, the general approach should be applicable to other HARDI acquisition schemes.

Julien Cohen-Adad¹, M Dylan Tisdall¹, Ralph Kimmlingen², Eva Eberlein², Thomas Witzel¹, Philipp Hoecht², Boris Keil¹, Juergen Nistler², Dietmar Lehne², Keith Heberlein³, Jennifer A McNab¹, Herbert Thein², Franz Schmitt², Bruce R Rosen¹, Van J Wedeen¹, and Lawrence L Wald^1,4
1A. A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA, United States, ²Siemens Healthcare, Erlangen, Germany, ³Siemens Healthcare, Boston, United States, ⁴Harvard-MIT Division of Health Sciences and Technology, MIT, Cambridge, MA, United States

In this work we used a novel Connectom Gradient (AS302) system specifically designed to achieve ultra-high gradient strength (Gmax = 300 mT/m , about 7.5 fold stronger than clinical scanners). This enables to shorten the diffusion-encoding time and thereby decrease the TE, yielding significant gains in SNR. In addition, we developed a 64-channel receive coil to further increase the SNR. We compare the in vivo HARDI data with Q-Ball reconstruction at variable gradient strengths and b-values up to 15000 s/mm2 in human. The higher SNR is shown to provide improvements in ODF metrics such as the fiber uncertainty and FA.

Aymeric Stamm¹, Patrick Pérez², and Christian Barillot^1
1Visages INSERM/INRIA U746, IRISA - UMR CNRS 6074, Rennes, France, ²Technicolor, Rennes, France

Clinical brain diffusion imaging provides low angular resolution diffusion images that were up to now analyzed under the assumption of Gaussian diffusion. An important limitation of such an analysis is that it does not exhibit crossing fibers. In this abstract, a non-Gaussian diffusion profile is assumed that allows for the accurate estimation of crossing fibers from low angular resolution diffusion images.