EE Seminar: Handling huge fiber-sets in Diffusion-Weighted MRI Brain Analysis: The Fiber-Density-Coreset for redundancy reduction
Speaker: Guy Alexandroni,
M.Sc. student under the supervision of Prof. Hayit Greenspan.
Wednesday, January 11th, 2017 at 15:00
Room 011, Kitot Bldg., Faculty of Engineering
Handling huge fiber-sets in Diffusion-Weighted MRI Brain Analysis: The Fiber-Density-Coreset for redundancy reduction
Abstract
State of the art Diffusion Weighted Magnetic Resonance Imaging protocols (DW-MRI), followed by advanced tractography techniques, produce impressive reconstructions of white matter pathways in the brain. These pathways often contain millions of trajectories (fibers). While for several applications the high number of fibers is essential, other applications (visualization, registration, some types of across-subject comparison) can achieve satisfying results using much smaller sets and may be overburdened by the computational load of the large fiber sets.
In this work, we present a novel, highly efficient algorithm for extracting a meaningful subset of fibers, which we term the Fiber-Density-Coreset (FDC). The reduced set is optimized to represent the main structures of the brain. FDC is based on an efficient geometric approximation paradigm named coresets, an optimization scheme showing much success in tasks requiring large computation time and/or memory. The reduced sets were evaluated by several methods, including a novel structural comparison to the full sets called 3D indicator structure comparison. The comparison was applied to High Angular Resolution Diffusion Imaging (HARDI) scans of 15 healthy individuals obtained from the Human Connectome Project. FDC produced the most satisfying subsets, consistently in all subjects. It also displayed low memory usage and significantly lower running time than conventional fiber reduction schemes.
Additional tools, developed for handling huge fiber-sets such as fiber compression by sparse representation and automatic tract segmentation methods, will be briefly reviewed as well
