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

15 Oct 2020  ·  Zaccharie Ramzi, Philippe Ciuciu, Jean-Luc Starck ·

We present a new neural network, the XPDNet, for MRI reconstruction from periodically under-sampled multi-coil data. We inform the design of this network by taking best practices from MRI reconstruction 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.

PDF Abstract


Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Benchmark
MRI Reconstruction fastMRI Brain 4x XPDNet SSIM 0.9581 # 2
PSNR 41.3 # 1
MRI Reconstruction fastMRI Brain 8x XPDNet SSIM 0.9408 # 2
PSNR 38.1 # 1
MRI Reconstruction fastMRI Knee 4x XPDNet SSIM 0.9287 # 2
PSNR 40.2 # 1
MRI Reconstruction fastMRI Knee 8x XPDNet SSIM 0.8893 # 4
PSNR 37.2 # 2


No methods listed for this paper. Add relevant methods here