SCAMPI – database-free neural network reconstruction for undersampled Magnetic Resonance Imaging

Abstract Reducing acquisition time in Magnetic Resonance Imaging (MRI) is mainly based on vast undersampling of the data measured in spatial frequency domain. The resulting artifacts are reduced by various reconstruction algorithms. Here, we present SCAMPI Sparsity Constrained Application of deep Magnetic resonance Priors for Image reconstruction), a deep Neural Network, that utilizes the Deep Image Prior approach together with sparsity constraints to reconstruct undersampled MRI data without any previous training on external datasets. Two-dimensional MRI data from the public FastMRI dataset with Cartesian undersampling in phase-encoding direction were reconstructed for different acceleration rates for single coil and multicoil data. Performance of the proposed architecture was compared to Parallel Imaging with Compressed Sensing. SCAMPI outperforms these state-of-the-art algorithms by better reducing undersampling artifacts and yielding lower error metrics in multicoil imaging. It is a novel tool for reconstructing undersampled single coil k-space data. Furthermore, SCAMPI works on multicoil data without explicit knowledge of coil sensitivity profiles. The presented approach avoids overfitting to dataset features, that can occur in Neural Networks trained on databases.