CNR-based Variable Resolution Reconstruction for Black Blood MRA
Vemuri P, Parker D, Kholmovski E
University of Utah, University of Utah
In present day MR, faster imaging strategies facilitate imaging with higher spatial resolution. But since noise in MR data is uniformly distributed throughout k-space, an increase in the amount of k-space sampled for increasing the spatial resolution results in the addition of more noise to the image. This abstract derives a theoretical dependence of the CNR of vessels in black-blood MRA on the fraction of k-space used for reconstruction. Based on this analysis, we present a variable resolution reconstruction technique which utilizes the local CNR value to improve the visibility of the vascular structures in black blood MRA.