Samuel Fielden1,
Alexander Klibanov1, 2, Frederick H. Epstein1,
3, Yaqin Xu1, Brent French1, 2, Craig
Meyer1, 3
1Biomedical
Engineering, University of Virginia, Charlottesville, VA, United States; 2Medicine,
University of Virginia, Charlottesville, VA, United States; 3Radiology,
University of Virginia, Charlottesville, VA, United States
Fluorine-19 (19F) is an attractive second-nucleus for contrast-enhanced MRI. However, for 19F cellular and molecular imaging, in order to achieve useful image resolution in reasonable scan times, methods must be developed to efficiently generate and utilize signal arising from small volumes of contrast agent. Here, we attempt to improve 19F imaging of cardiac inflammation through a 3-pronged approach: First, through pulse sequences optimized to maximize SNR; second, by incorporating side information from proton images in a constrained iterative reconstruction; finally, by exploiting the underlying sparsity of 19F images using model-based image reconstructions.