Constrained Reconstruction for Sparse Magnetic Resonance Imaging

Abstract

Magnetic Resonance Imaging (MRI) often requires acquisition time reduction to measure dynamic processes changes. To this aim is necessary to reduce the number of measured data. This results in an undersampling problem and aliasing. In what follows, a simple constrained reconstruction algorithm for sparse k-space sampling is described, having the scope of reducing the undersampling artefacts. The proposed method can be applied to different k-space trajectories. Its performance has been demonstrated on MRI data sampled numerically using different trajectories. The presented method has been also compared with other interpolation techniques and results are reported.