Highly Constrained Back Projection (HYPR): Theory and Potential MRI Applications
Block W, Velikina J, Wu Y, Johnson K, Korosec F, Mistretta C, Perry J, Wu Y, Wieben O
University of Wisconsin
This study investigates a novel, non-iterative, unfiltered backprojection algorithm that effectively allows for the acquisition and reconstruction of a time series with higher temporal resolution while maintaining high SNR and a low artifact level. The technique was evaluated in simulations by synthesizing numerical phantoms for fMRI and MR spectroscopy and by generating undersampled 2D images obtained from contrast-enhanced MRA, breast perfusion, diffusion, and CT perfusion imaging to investigate artifacts and waveform fidelity. Subsequently, HYPR data were acquired in phantom studies and in CE-MRA using PR TRICKS. Significant acquisition acceleration factors were possible in a wide range of applications.