Synopsis

This talk will review advanced reconstruction methods for DCE-MRI and DSC-MRI. These methods allow for superior extraction of perfusion, permeability, and other tracer-kinetic parameters of interest (from noisy data), and allow for improvements in spatio-temporal resolution and coverage through the use of sparse data sampling (in k,t space). We will start by reviewing the current workflow in DCE/DSC, and identifying the opportunity for improvement. Then we will go through four classes of reconstruction, which have some overlap: 1) spatial and temporal sparsity constraints, 2) data-driven constraints, 3) tracer-kinetic model-based constraints, 4) arterial input function estimation. Finally, we will discuss the magnitude of the improvement that can be achieved using these techniques.