To perform sparse-data CT, the iterative reconstruction commonly uses regularizers in the CS framework. In this paper, inspired by the idea of deep learning, we unfold a state-of-the-art “fields of experts” based iterative reconstruction scheme up to a number of iterations for data-driven training, construct a Learned Experts’ Assessment-based Reconstruction Network (“LEARN”) for sparse-data CT, and demonstrate the feasibility and merits of LEARN network. The experimental results produces a superior performance to several state-of-the-art methods.

This presentation is part of Minisymposium “MS24 - Data-driven approaches in imaging science (3 parts)”
organized by: Jae Kyu Choi (Institute of Natural Sciences, Shanghai Jiao Tong University) , Chenglong Bao (Yau Mathematical Sciences Center, Tsinghua University) .

Authors:

Yi Zhang (School of Computer Science, Sichuan University)

Hu Chen (Sichuan University)

Yunjin Chen (ULSee Inc.)

Weihua Zhang (Sichuan University)

Huaiqiang Sun (West China Hospital of Sichuan University)

Yang Lv (Shanghai United Imaging Healthcare Co., Ltd)

Peixi Liao (The Sixth People’s Hospital of Chengdu)

Ge Wang (Rensselaer Polytechnic Institute)

Keywords:

computed tomography, deep learning, image reconstruction, inverse problems, machine learning