XPDNet for MRI Reconstruction: an application to the 2020 fastMRI challenge

We present a new neural network, the XPDNet, for <PRE_TAG>MRI reconstruction</POST_TAG> from periodically under-sampled multi-coil data. We inform the design of this network by taking best practices from <PRE_TAG>MRI reconstruction</POST_TAG> and computer vision. We show that this network can achieve state-of-the-art reconstruction results, as shown by its ranking of second in the fastMRI 2020 challenge.