LEARN: Learned Experts’ Assessment-based Reconstruction Network (LEARN) for Sparse-Data CT

If you use this package in your research, please cite the following paper:

Hu Chen, Yi Zhang, Yunjin Chen, Junfeng Zhang, Weihua Zhang, Huaiqiang Sun, Yang Lv, Peixi Liao, Jiliu Zhou and Ge Wang, "LEARN: Learned Experts’ Assessment-based Reconstruction Network for Sparse-data CT" IEEE Trasnactions on Medical Imaging, 37(6), 1333-1347. 2018. This code includes the MATLAB file of LEARN and the demo code for verifying it using MatConvnet. GPU(cuda) should be installed.

Training: GenerateData_model.m: Generate the train data and test data. Demo_Train_model.m: Code for training.

Testing: Demo_Test_model.m: Code for testing. LEARN_MODEL.mat: the trained model. testimg.mat:the image for testing. proMatrix_64.mat:projection matrix

Please Contact:
Hu Chen huchen@scu.edu.cn or Yi Zhang yzhang@scu.edu.cn If you have any questions about it.