Meeting Banner
Abstract #3860

McKinnon-Bates Sparsification for Accelerated Phase Contrast Imaging

Andrew L. Wentland1, 2, Oliver Wieben1, Thomas M. Grist2, Kevin M. Johnson1

1Medical Physics, University of Wisconsin School of Medicine & Public Health, Madison, WI, United States; 2Radiology, University of Wisconsin School of Medicine & Public Health, Madison, WI, United States

Numerous procedures have been proposed in the literature to accelerate the acquisition of phase contrast (PC) data. However, these techniques require substantial reconstruction times and can have unpredictable artifacts and/or performance. We have explored sparsifying preprocessing techniques to provide relatively quick reconstruction of 2DPC data with very few projections. Initial results produced magnitude and velocity images of good quality with acceleration factors up to 30. Flow measurements in the processed images had good temporal fidelity and were consistent with measurements from unprocessed images. This algorithm may be useful for reducing the acquisition time of 2DPC data or improving temporal resolution.

Keywords

accelerated acceleration achieved allow allowed aorta aortic apply artery artifacts ascending background bates blast blurring breath cardiac challenging characteristics coils combined composite compressed contrast corr correlation created cross curves cycle dark depends despite deteriorate effective encouraging especially even expected experiences feasible flow frame frames free gating good grist health highly hold improve improvements improving investigate lead localized magnitude marked mask measured medical medicine nevertheless noise obscure oliver original parallel parsed past patients patterns peak performance physics plotted preprocessing preserved previous probability process processed processing produces projections proposed pseudo public pulmonary quality radial radially radiology random reconstructed reconstruction reconstructions reduced reducing reductions relatively reliable relies represent representative require resolution retrospective robust sacrificed sampled sampling scanner schematic school sensing sensitivities shaded slice sliding sorted sparsity spatial static still streak substantial substantially subtract subtracted success suffer sufficient summarized systolic temporal temporally theory tissue trajectories trans unpredictable unprocessed useful utilized utilizing variable velocity vessels window