Sajan Goud Lingala1, Yue Hu2,
Edward Dibella3, Mathews Jacob1
1Biomedical Engineering,
University of Rochester, Rochester, NY, United States; 2Electrical
& Computer Engineering, University of Rochester, Rochester, NY, United
States; 3Radiology, University of Utah, Salt Lake city, UT, United
States
We consider the problem of minimizing the trade-offs between the image parameters (spatio-temporal resolution, volume coverage and the SNR) routinely observed in myocardial perfusion MRI. In this context, we propose to use our accelerated k-t SLR scheme (which exploits the low rank and sparsity properties of the dynamic data) in combination with parallel imaging. Experimental results and comparisons show that, our proposed scheme provide superior reconstructions with better fidelity at high accelerations (>10), while existing schemes such as k-t PCA, k-t FOCUSS and TV based spatio-temporal regularizers have limitations characteristic to their method of operation