Low-resolution Prior Equilibrium Network for CT Reconstruction (2401.15663v2)
Abstract: The unrolling method has been investigated for learning variational models in X-ray computed tomography. However, it has been observed that directly unrolling the regularization model through gradient descent does not produce satisfactory results. In this paper, we present a novel deep learning-based CT reconstruction model, where the low-resolution image is introduced to obtain an effective regularization term for improving the network`s robustness. Our approach involves constructing the backbone network architecture by algorithm unrolling that is realized using the deep equilibrium architecture. We theoretically discuss the convergence of the proposed low-resolution prior equilibrium model and provide the conditions to guarantee convergence. Experimental results on both sparse-view and limited-angle reconstruction problems are provided, demonstrating that our end-to-end low-resolution prior equilibrium model outperforms other state-of-the-art methods in terms of noise reduction, contrast-to-noise ratio, and preservation of edge details.
- Zhao S, Xia D and Zhao X 2021 Inverse Problems 37 075001
- Kang E, Min J and Ye J C 2017 Medical Physics 44 e360–e375
- Zickert G, Öktem O and Yarman C E 2022 Inverse Problems 38 105010
- Gholizadeh-Ansari M, Alirezaie J and Babyn P 2020 Journal of Digital Imaging 33 504–515
- Baguer D O, Leuschner J and Schmidt M 2020 Inverse Problems 36 094004
- Schwab J, Antholzer S and Haltmeier M 2019 Inverse Problems 35 025008
- Lunz S, Öktem O and Schönlieb C B 2018 Advances in Neural Information Processing Systems 31
- Chan S H, Wang X and Elgendy O A 2016 IEEE Transactions on Computational Imaging 3 84–98
- Nair P, Gavaskar R G and Chaudhury K N 2021 IEEE Transactions on Computational Imaging 7 337–348
- Bai S, Kolter J Z and Koltun V 2019 Advances in Neural Information Processing Systems 32
- Gilton D, Ongie G and Willett R 2021 IEEE Transactions on Computational Imaging 7 1123–1133
- Venkatakrishnan S V, Bouman C A and Wohlberg B 2013 Plug-and-play priors for model based reconstruction 2013 IEEE global conference on signal and information processing (IEEE) pp 945–948
- Hurault S, Leclaire A and Papadakis N 2021 arXiv preprint arXiv:2110.03220
- Monga V, Li Y and Eldar Y C 2021 IEEE Signal Processing Magazine 38 18–44
- Gregor K and LeCun Y 2010 Learning fast approximations of sparse coding Proceedings of the 27th international conference on international conference on machine learning pp 399–406
- Bard J F 2013 Practical bilevel optimization: algorithms and applications vol 30 (Springer Science & Business Media)
- Bracken J and McGill J T 1973 Operations Research 21 37–44
- Colson B, Marcotte P and Savard G 2007 Annals of Operations Research 153 235–256
- Engl H W, Hanke M and Neubauer A 1996 Regularization of inverse problems vol 375 (Springer Science & Business Media)
- Dempe S and Zemkoho A 2020 Springer optimization and its applications 161
- Ghadimi S and Wang M 2018 arXiv preprint arXiv:1802.02246
- Fessler J 2014
- Censor Y 1983 Proceedings of the IEEE 71 409–419
- Hansen P C 1998 Rank-deficient and discrete ill-posed problems: numerical aspects of linear inversion (SIAM)
- Stearns C W, Manjeshwar R M and Wollenweber S D 2006 An efficient algorithm for targeted reconstruction of tomographic data 2006 IEEE Nuclear Science Symposium Conference Record vol 5 (IEEE) pp 2808–2811
- Dabravolski A, Batenburg K J and Sijbers J 2014 PloS One 9 e106090
- Malik V and Hussein E M 2021 Heliyon 7 e06839
- Adler J and Öktem O 2018 IEEE Transactions on Medical Imaging 37 1322–1332
- Yuan H, Jia J and Zhu Z 2018 Sipid: A deep learning framework for sinogram interpolation and image denoising in low-dose ct reconstruction 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018) (IEEE) pp 1521–1524
- McCollough C 2016 Medical Physics 43 3759–3760