Vihang Agarwal1, Yue Cao1, and James Balter1
1Radiation Oncology, University of Michigan, Ann Arbor, MI, United States
1Radiation Oncology, University of Michigan, Ann Arbor, MI, United States
We demonstrate joint optimization of under-sampling patterns with the proposed image reconstruction neural network architecture allows under-sampling with higher acceleration rates and produces superior quality images as compared to random under-sampling schemes.
The architecture of our Attentive Residual Non Local-Network (ARNL-Net) with RCAB as building blocks.
Image reconstructions under cartesian constrained acquisition scheme. Second row demonstrates the results for the joint optimization model.
