ROVir Enables Substantially Easier Real-Time Imaging of Small Regions-of-Interest

Motivation: Many real-time MRI experiments use large FOVs, even though most of the FOV is uninformative. Using a large FOV helps avoid aliasing but requires a higher level of undersampling and more complicated reconstruction. The ROVir (region-optimized virtual coils) technique can potentially alleviate these issues. Goal(s): To develop a more efficient approach to real-time imaging of small regions-of-interest. Approach: We use ROVir to suppress the signal from uninteresting spatial regions, enabling a smaller FOV (with closer-to-Nyquist k-space sampling). This allows substantially simpler reconstruction. Results: ROVir with simple reconstruction achieves similar/potentially-better image quality than more sophisticated regularization-based reconstructions, with substantially less computation. Impact: High-resolution real-time MRI over a large FOV is usually highly-undersampled and requires advanced/time-consuming reconstruction methods. We demonstrate that ROVir can be used to shrink the size of the FOV to enable a much simpler/easier reconstruction problem.