Meeting Banner
Abstract #4174

How Well Can Magnetic Susceptibility Anisotropy Be Estimated? an Error Analysis of Cylindrically Symmetric Susceptibility Tensor Reconstructions from Few Orientations

Cynthia Wisnieff1, Pascal Spincemaille2, Tian Liu3, Yi Wang1

1Cornell Univerisity, New York, United States; 2Weill Cornell Medical College, New York, United States; 3Medimagemetric, New York, United States

Susceptibility tensor imaging, STI, has shown evidence of being intrinsically linked to the organization and composition of the myelin sheath of white matter. However, estimating the magnetic susceptibility anisotropy within human subjects is limited by both the acquisition time necessary and the difficulty in reorienting the subject in the scanner. In this work we present an error analysis of constrained system used for cylindrically symmetric susceptibility tensor imaging reconstructions and examine the feasibility of estimating the anisotropy with these constrained reconstructions with few subject orientations in simulations and in vivo.

Abstract PDF Presentation Video

Keywords

able acquiring acquisition acquisitions additional agreeable aligned allowed alternative amplification anisotropy anterior anyone applying approval arrows audience black body bottom brain circles clinical college color combinations computed condition consistent consistently constrained constraints constructed corpus correspond cylindrically dependence diamonds difficulty diffusion distributed echoes error errors estimating estimation exactly examination feasible fewer fiber fibers field frame gradient hampered head highly human imposed in vivo indicated indicator inferior infinite inverse isotropic making many maps matrix measured median medical natl necessary numerical optic organization orientation orientations pascal phantom pitfalls positions posterior potential pursued radiations realistic reconstruct reconstructed reconstruction reconstructions reduce reliable remaining report represent resolution respectively rotation sampled scanner sensitive sets simulated solution solving space spacing sphere subject subjects substantial superior susceptibility symmetric table target tensor tilts underestimation uniform uniformly unknowns various volunteer volunteers white worst