ABSTRACT
In what follows, some interesting methods for sparse sampling (undersampling) acquisition and reconstruction are presented. In particular, a differentiation is made between those reconstruction/restoration methods that modify the k-space trajectories during acquisition, that is, the chosen trajectories (both in number and directions) depend on the sample shape, and those methods that reduce artifacts independently of the sample shape. The first class contains adaptive methods, the second includes compressed sensing. As a further examination, the possibility of jointly using adaptive methods in a compressed sensing strategy has been also explored and some preliminary results, demonstrating that these joint methods can allow a further reduction of the dimension of the data set necessary for good reconstruction, are reported. The implementation and practical impact of sparse sampling methods are also discussed.
