Abstract #3656

Lin F

An auto-calibration scan (ACS) can improve parallel MRI reconstructions, including the popular GRAPPA method. Usually, the ACS is collected at the center of k-space, constituting low-resolution prior information. Previously, we have reported the advantage of prior-informed SENSE reconstruction with reduced noise amplification. Here we propose a prior-regularized GRAPPA reconstruction to further stabilize the reconstructed images at high acceleration rates. We demonstrated prior-informed GRAPPA on a 3T high-resolution anatomical scan. This method can be generalized for optimized k-t space sampling to provide priors with identical spatial resolution.