PWLS-ULTRA: An Efficient Clustering and Learning-Based Approach for Low-Dose 3D CT Image Reconstruction
IEEE Xplore: https://ieeexplore.ieee.org/document/8352798/
Arxiv: https://arxiv.org/abs/1703.09165
We provide XCAT phantom [1] (both 2D slices and 3D volume), pre-learned transforms (and dictionary), and reconstructions for initialization, which were used in this paper. So you could quickly reproduce results in our paper right after running the main code of one specific method.
Codes should be run with Michigan Image Reconstruction Toolbox (MIRT). http://web.eecs.umich.edu/~fessler/code/index.html
Note that this toolbox only supports Linux/Mac. Before running the main code for reconstruction (e.g., "main_axial_pwls_ultra.m/main_fan_pwls_ultra.m"), you just need to run "setup.m" in MIRT. And then run "axial_proj_data_maker_HU.m/fan_proj_data_maker_HU.m" to pre-compute the (noisy) sinograms, weighting matrix, kappa map, and majorzing matrix, which are stored in "data/3Dxcat/tmp/" or "data/2Dxcat/tmp/".
We also provide our own implementation for the methods compared in this paper: PWLS-EP [2], PWLS-DL [3].
[1] W. P. Segars, M. Mahesh, T. J. Beck, E. C. Frey, and B. M. W. Tsui, “Realistic CT simulation using the 4D XCAT phantom,” Med. Phys., vol. 35, no. 8, pp. 3800–3808, Aug. 2008.
[2] H. Nien and J. A. Fessler, “Relaxed linearized algorithms for faster X-ray CT image reconstruction,” IEEE Trans. Med. Imag., vol. 35, no. 4, pp. 1090–1098, Apr. 2016.
[3] Q. Xu, H. Yu, X. Mou, L. Zhang, J. Hsieh, and G. Wang, “Low-dose X-ray CT reconstruction via dictionary learning,” IEEE Trans. Med. Imag., vol. 31, no. 9, pp. 1682–1697, Sep. 2012.
If you have any questions or suggestions, please feel free to contact me via [email protected].