Doubly Adaptive Estimation of Dynamic Mri Sequences

Dynamic magnetic resonance imaging (MRI) refers to the repeated acquisition of image data to monitor changes in tissue structure over time. In this paper we present a method for the efficient acquisition of dynamic MRI sequences based on two complimentary strategies: an adaptive framework for the estimation of the MRI images themselves and a method to dynamically adapt the MRI system excitations. Thus, we denote this method as doubly adaptive. Analysis shows that the previously proposed Fourier Keyhole and SVD methods are specific cases of the general adaptive framework presented here. We also show that complete knowledge of the next image in the sequence is required to determine the optimal system inputs. Since this knowledge is not realizable, we introduce a linear predictor to aid in determining appropriate system inputs. A simulated example using real MRI data is given showing that the doubly adaptive strategy can provide significant improvement over current minimal MRI data acquisition and image estimation techniques.