Article Contents
Article Contents

A nonconvex truncated regularization and box-constrained model for CT reconstruction

• * Corresponding author: Yiming Gao
This work was supported by Zhejiang Provincial Natural Science Foundation of China (No. LQ20A010007) the National Natural Science Foundation of China (NSFC) (Nos. 11871035, 11531013, 11971138), Recruitment Program of Global Young Expert, and Postdoctoral Science Foundation of China (No. 2019M651002)
• X-ray computed tomography has been a useful technology in cancer detection and radiation therapy. However, high radiation dose during CT scans may increase the underlying risk of healthy organs. Usually, sparse-view X-ray projection is an effective method to reduce radiation. In this paper, we propose a constrained nonconvex truncated regularization model for this low-dose CT reconstruction. It preserves sharp edges very well. Although this model is quite complicated to analyze, we establish two useful theoretical results for its minimizers. Motivated by them, an iterative support shrinking algorithm is introduced. To handle more nondifferentiable points of the regularization function except zero point, we use a general proximally linearization technique at them, which is helpful to implement our algorithm. For this algorithm, we prove the convergence of the objective function, and give a lower bound theory of the iterative sequence. Numerical experiments and comparisons demonstrate that our model with the proposed algorithm performs good for low-dose CT reconstruction.

Mathematics Subject Classification: Primary: 68U10, 65K10, 94A08, 90C26.

 Citation:

• Figure 1.  Three different $\varphi(t)$. (a) $\varphi(t) = t,\ 0 \leq t \leq 0.5; \ \varphi(t) = 0.5,\ t > 0.5$. (b) $\varphi(t) = t^{0.5},\ 0 \leq t \leq 0.5; \ \varphi(t) = 0.5^{0.5},\ t > 0.5$. (c) $\varphi(t) = \mathrm{log}(t+1),\ 0 \leq t \leq 1; \ \varphi(t) = \mathrm{log}(t^{0.5}+1),\ 1 < t \leq 2; \ \varphi(t) = \mathrm{log}(2^{0.5}+1),\ t >2$

Figure 2.  Test images. (a) $128\times 128$; (b) $256\times 256$

Figure 3.  The visualization of PSNR values versus parameters $a$ and $\tau$ of trunc-LN, $p$ and $\tau$ of trunc-$\ell_p$ for "Shepp-Logan'' image with 36 projections

Figure 4.  CT reconstruction comparisons for "Shepp-Logan'' image with $\sigma = 0.01\|g\|_\infty$. The first, second and third rows are the reconstructed results when $N = 18$, 36 and 72, respectively. The PSNR and SSIM values are attached in the brackets

Figure 5.  CT reconstruction comparisons for "NCAT'' image with $\sigma = 0.01\|g\|_\infty$. The first, second and third rows are the reconstructed results when $N = 18$, 36 and 72, respectively. The PSNR and SSIM values are attached in the brackets

Figure 6.  The first row shows the ground truth. The second and third rows show the zoom-in views of the "Shepp-Logan'' and "NCAT'' image reconstructions with 36 projections

Figure 7.  The residual errors of the reconstructions by three methods with 36 projections for "Shepp-Logan'' and "NCAT''

Figure 8.  The reconstruction comparisons of 60th and 80th rows of "Shepp-Logan'' image for the three methods with 36 projections

Figure 9.  The values $F(u_k)$ versus the iteration number for "Shepp-Logan'' image with 36 projections

Figure 10.  CT reconstruction comparisons for real chest CT image with projection number $N = 60$. The PSNR and SSIM values are attached in the brackets

Figure 11.  A simple example. (a) The illustration of an image and its gradient. (b) Pixel values in black and norms of gradient in red

Table 1.  More quantitative comparisons of reconstruction on "Shepp-Logan'' in terms of PSNR and SSIM

 $N_p$ noise level TV TW-$\ell_0$ trunc-LN trunc-$\ell_p$ PSNR/SSIM PSNR/SSIM PSNR/SSIM PSNR/SSIM $18$ $0.005\|g\|_\infty$ 36.28/0.9688 39.31/0.9871 43.36/0.9902 43.66/0.9924 $0.01\|g\|_\infty$ 31.88/0.9159 34.67/0.9729 39.19/0.9800 38.24/0.9794 $0.02\|g\|_\infty$ 27.52/0.8389 29.54/0.9460 32.08/0.9507 31.53/0.9561 $36$ $0.005\|g\|_\infty$ 39.49/0.9860 44.52/0.9941 46.99/0.9959 49.52/0.9991 $0.01\|g\|_\infty$ 34.45/0.9564 39.84/0.9860 42.80/0.9897 42.87/0.9910 $0.02\|g\|_\infty$ 29.71/0.8972 33.56/0.9695 38.09/0.9776 37.43/0.9769 $72$ $0.005\|g\|_\infty$ 43.27/0.9916 49.93/0.9984 50.20/0.9980 52.92/0.9989 $0.01\|g\|_\infty$ 37.80/0.9779 43.95/0.9942 45.61/0.9954 46.41/0.9954 $0.02\|g\|_\infty$ 32.54/0.9393 37.36/0.9758 41.07/0.9862 40.58/0.9868

Table 2.  More quantitative comparisons of reconstruction on "NCAT'' in terms of PSNR and SSIM

 $N_p$ noise level TV TW-$\ell_0$ trunc-LN trunc-$\ell_p$ PSNR/SSIM PSNR/SSIM PSNR/SSIM PSNR/SSIM $18$ $0.005\|g\|_\infty$ 30.89/0.9580 29.97/0.9704 33.27/0.9875 33.90/0.9887 $0.01\|g\|_\infty$ 28.15/0.9141 26.74/0.9399 28.52/0.9630 28.44/0.9658 $0.02\|g\|_\infty$ 25.39/0.8220 24.12/0.8956 24.69/0.9189 24.72/0.9216 $36$ $0.005\|g\|_\infty$ 33.49/0.9703 34.01/0.9877 38.36/0.9954 38.99/0.9964 $0.01\|g\|_\infty$ 30.20/0.9339 29.48/0.9668 31.32/0.9816 31.79/0.9834 $0.02\|g\|_\infty$ 27.08/0.8448 25.88/0.9210 27.02/0.9520 27.19/0.9548 $72$ $0.005\|g\|_\infty$ 35.78/0.9721 37.81/0.9943 42.29/0.9980 42.76/0.9985 $0.01\|g\|_\infty$ 31.70/0.9340 31.86/0.9795 34.25/0.9902 34.40/0.9908 $0.02\|g\|_\infty$ 27.89/0.8276 27.44/0.9410 28.94/0.9694 29.18/0.9714
•  [1] A. H. Andersen and A. C. Kak, Simultaneous algebraic reconstruction technique (SART): A superior implementation of the ART algorithm, Ultrasonic Imaging, 6 (1984), 81-94. [2] C. Bao, B. Dong, L. Hou, Z. Shen, X. Zhang and X. Zhang, Image restoration by minimizing zero norm of wavelet frame coefficients, Inverse Problems, 32 (2016), 115-142.  doi: 10.1088/0266-5611/32/11/115004. [3] M. Bhatia, W. C. Karl and A. S. Willsky, A wavelet-based method for multiscale tomographic reconstruction, IEEE Transactions on Medical Imaging, 15 (1996), 92-101.  doi: 10.21236/ADA459987. [4] E. G. Birgin, J. M. Martínez and M. Raydan, Nonmonotone spectral projected gradient methods on convex sets, SIAM Journal on Optimization, 10 (2000), 1196-1211.  doi: 10.1137/S1052623497330963. [5] D. J. Brenner and E. J. Hall, Computed Tomography–an increasing source of radiation exposure, New England Journal of Medicine, 357 (2007), 2277-2284.  doi: 10.1056/NEJMra072149. [6] A. Cai, L. Wang, B. Yan, L. Li, H. Zhang and G. Hu, Efficient TpV minimization for circular, cone-beam computed tomography reconstruction via non-convex optimization, Computerized Medical Imaging and Graphics, 45 (2015), 1-10.  doi: 10.1016/j.compmedimag.2015.06.004. [7] R. H. Chan, M. Tao and X. Yuan, Constrained total variation deblurring models and fast algorithms based on alternating direction method of multipliers, SIAM Journal on Imaging Sciences, 6 (2013), 680-697.  doi: 10.1137/110860185. [8] X. Chen, Smoothing methods for nonsmooth, nonconvex minimization, Mathematical Programming, 134 (2012), 71-99.  doi: 10.1007/s10107-012-0569-0. [9] X. Chen, M. K. Ng and C. Zhang, Non-Lipschitz $\ell_{p}$-Regularization and Box Constrained Model for Image Restoration, IEEE Transactions on Image Processing, 21 (2012), 4709-4721.  doi: 10.1109/TIP.2012.2214051. [10] X. Chen, F. Xu and Y. Ye, Lower bound theory of nonzero entries in solutions of $\ell_{2}$-$\ell_{p}$ minimization, SIAM Journal on Scientific Computing, 32 (2010), 2832-2852.  doi: 10.1137/090761471. [11] Z. Chen, X. Jin, L. Li and G. Wang, A limited-angle CT reconstruction method based on anisotropic TV minimization, Physics in Medicine and Biology, 58 (2013), 2119-2141.  doi: 10.1088/0031-9155/58/7/2119. [12] J. K. Choi, B. Dong and X. Zhang, Limited tomography reconstruction via tight frame and simultaneous sinogram extrapolation, J. Comput. Math., 34 (2016), 575-589.  doi: 10.4208/jcm.1605-m2016-0535. [13] B. Dong, J. Li and Z. Shen, X-Ray CT image reconstruction via wavelet frame based regularization and radon domain inpainting, J. Sci. Comput., 54 (2013), 333-349.  doi: 10.1007/s10915-012-9579-6. [14] G. Edgar and T. H. Herman, X-Ray CT image reconstruction via wavelet frame based regularization and radon domain inpainting, Inverse Problems, 33 (2017), 185-208. [15] A. J. Einstein, M. J. Henzlova and R. Sanjay, Estimating risk of cancer associated with radiation exposure from 64-slice computed tomography coronary angiography, Journal of American Medical Association, 298 (2007), 317-323.  doi: 10.1001/jama.298.3.317. [16] E. Y. Sidky, R. Chartrand, J. M. Boone and X. Pan, Constrained TpV minimization for enhanced exploitation of gradient sparsity: Application to CT image reconstruction, IEEE Journal of Translational Engineering in Health and Medicine, 2 (2014), 1-18. [17] H. Gao, R. Li, Y. Lin and L. Xing, 4D cone beam CT via spatiotemporal tensor framelet, Medical Physics, 39 (2012), 6943-6946.  doi: 10.1118/1.4762288. [18] Y. Gao and C. Wu, On a general smoothly truncated regularization for variational piecewise constant image restoration: Construction and convergent algorithms, Inverse Problems, 36 (2020), 1-31.  doi: 10.1088/1361-6420/ab6619. [19] R. Gordon, R. Bender and G. T. Herman, Algebraic reconstruction techniques (ART) for three-dimensional electron microscopy and x-ray photography, Journal of Theoretical Biology, 29 (1970), 471-476.  doi: 10.1016/0022-5193(70)90109-8. [20] M. Hintermüller and T. Wu, Nonconvex $TV^{q}$-models in image restoration: Analysis and a trust-region regularization based superlinearly convergent solver, SIAM Journal on Imaging Sciences, 6 (2013), 1385-1415.  doi: 10.1137/110854746. [21] X. Jia, B. Dong, Y. Lou and S. B. Jiang, GPU-based iterative cone-beam CT reconstruction using tight frame regularization, Physics in Medicine and Biology, 56 (2011), 3787-3801.  doi: 10.1088/0031-9155/56/13/004. [22] X. Jia, Y. Lou, R. Li, W. Y. William and S. B. Jiang, GPU-based fast cone beam CT reconstruction from undersampled and noisy projection data via total variation, Medical Physics, 37 (2010), 1757-1760.  doi: 10.1118/1.3371691. [23] M. Jiang and G. Wang, Convergence of the simultaneous algebraic reconstruction technique (SART), IEEE Transactions on Image Processing, 12 (2003), 957-961.  doi: 10.1109/TIP.2003.815295. [24] H. Kim, J. Chen, A. Wang, C. Chuang, M. Held and J. Pouliot, Non-local total-variation (NLTV) minimization combined with reweighted L1-norm for compressed sensing CT reconstruction, Physics in Medicine and Biology, 61 (2016), 6878-6891.  doi: 10.1088/0031-9155/61/18/6878. [25] J. K. Kim, J. A. Fessler and Z. Zhang, Forward-projection architecture for fast iterative image reconstruction in X-ray CT, IEEE Transactions on Signal Processing, 60 (2012), 5508-5518.  doi: 10.1109/TSP.2012.2208636. [26] J. W. Kress, L. A. Feldkamp and L. C. Davis, Practical cone-beam algorithm, Journal of the Optical Society of American, 1 (1984), 612-619. [27] M. Lai, Y. Xu and W. Yin, Improved iteratively reweighted least squares for unconstrained smoothed $\ell_{q}$ minimization, SIAM Journal on Numerical Analysis, 51 (2013), 927-957.  doi: 10.1137/110840364. [28] Y. Liu, Z. Liang, J. Ma, H. Lu, K. Wang, H. Zhang and W. Moore, Total variation-stokes strategy for sparse-view X-ray CT image reconstruction, IEEE Transactions on Medical Imaging, 33 (2014), 749-763. [29] Y. Liu, J. Ma, Y. Fan and Z. Liang, Adaptive-weighted total variation minimization for sparse data toward low-dose x-ray computed tomography image reconstruction, Physics in Medicine and Biology, 57 (2012), 7923-7956.  doi: 10.1088/0031-9155/57/23/7923. [30] M. Nikolova, Analysis of the Recovery of Edges in Images and Signals by Minimizing Nonconvex Regularized Least-Squares, SIAM Journal on Multiscale Modeling and Simulation, 4 (2005), 960-991.  doi: 10.1137/040619582. [31] M. Nikolova, M. K. Ng and C. P. Tam, Fast Nonconvex Nonsmooth Minimization Methods for Image Restoration and Reconstruction, IEEE Transactions on image processing, 19 (2010), 3073-3088.  doi: 10.1109/TIP.2010.2052275. [32] M. Nikolova, M. K. Ng, S. Zhang and W. K. Ching, Efficient reconstruction of piecewise constant images using nonsmooth nonconvex minimization, SIAM Journal on Imaging Sciences, 1 (2008), 2-25.  doi: 10.1137/070692285. [33] S. Niu, J. Huang, Z. Bian, D. Zeng, W. Chen, G. Yu, Z. Liang and J. Ma, Iterative reconstruction for sparse-view x-ray CT using alpha-divergence constrained total generalized variation minimization, Journal of X-ray Science and Technology, 25 (2017), 673-688.  doi: 10.3233/XST-16239. [34] S. Ramani and J. A. Fessler, A splitting-based iterative algorithm for accelerated statistical X-ray CT reconstruction, IEEE Transactions on Medical Imaging, 31 (2012), 677-688.  doi: 10.1109/TMI.2011.2175233. [35] M. Rantala, S. Vanska, S. Jarvenpaa, M. Kalke, M. Lassas, J. Moberg and S. Siltanen, Wavelet-based reconstruction for limited-angle X-ray tomography, IEEE Transactions on Medical Imaging, 25 (2006), 210-217.  doi: 10.1109/TMI.2005.862206. [36] L. I. Rudin, S. Osher and E. Fatemi, Nonlinear total variation based noise removal algorithms, Physica D: Nonlinear Phenomena, 60 (1992), 259-268.  doi: 10.1016/0167-2789(92)90242-F. [37] W. P. Segars, M. Mahesh, T. J. Beck, E. C. Frey and B. M. Tsui, Realistic CT simulation using the 4D XCAT phantom, Medical Physics, 35 (2008), 3800-3808.  doi: 10.1118/1.2955743. [38] E. Y. Sidky and X. Pan, Image reconstruction in circular cone-beam CT by constrained, total-variation minimization, Physics in Medicine and Biology, 153 (2008), 4777-4807. [39] J. Wang, T. Li and L. Xing, Iterative image reconstruction for CBCT using edge-preserving prior, Medical Physics, 36 (2009), 252-260.  doi: 10.1118/1.3036112. [40] W. Wang, C. Wu and X. Tai, A globally convergent algorithm for a constrained non-Lipschitz image restoration model, Journal of Scientific Computing, (2020). doi: 10.1007/s10915-020-01190-4. [41] Z. Wang, A. C. Bovik, H. R. Sheikh and E. P. Simoncelli, Image quality assessment: From error visibility to structural similarity, IEEE Transactions on Image Processing, 13 (2004), 600-612.  doi: 10.1109/TIP.2003.819861. [42] C. Wu, Z. Liu and S. Wen, A general truncated regularization framework for contrast-preserving variational signal and image restoration: Motivation and implementation, Science China Mathematics, 61 (2018), 1711-1732.  doi: 10.1007/s11425-017-9260-8. [43] C. Wu and X. Tai, Augmented Lagrangian method, dual methods, and split Bregman iteration for ROF, vectorial TV, and high order models, SIAM J. Imaging Sci., 3 (2010), 300-339.  doi: 10.1137/090767558. [44] Q. Xu, H. Yu, X. Mou, L. Zhang, H. Jiang and G. Wang, Low-dose X-ray CT reconstruction via dictionary learning, IEEE Transactions on Medical Imaging, 31 (2012), 1682-1697. [45] Y. Xu and W. Yin, A block coordinate descent method for regularized multiconvex optimization with applications to nonnegative tensor factorization and completion, SIAM Journal on Imaging Sciences, 6 (2013), 1758-1789.  doi: 10.1137/120887795. [46] C. Zeng, R. Jia and C. Wu, An iterative support shrinking algorithm for non-Lipschitz optimization in image restoration, Journal of Mathematical Imaging and Vision, 61 (2018), 122-139.  doi: 10.1007/s10851-018-0830-0. [47] C. Zeng and C. Wu, On the discontinuity of images recovered by noncovex nonsmooth regularized isotropic models with box constraints, Advances in Computational Mathematics, 45 (2018), 589-610.  doi: 10.1007/s10444-018-9629-1. [48] C. Zeng and C. Wu, On the edge recovery property of noncovex nonsmooth regularization in image restoration, SIAM Journal Numerical Analysis, 56 (2018), 1168-1182.  doi: 10.1137/17M1123687. [49] R. Zhan and B. Dong, CT image reconstruction by spatial-radon domain data-driven tight frame regularization, SIAM Journal on Imaging Sciences, 9 (2016), 1063-1083.  doi: 10.1137/16M105928X. [50] X. Zhang, M. Bai and M. K. Ng, Nonconvex-TV based image restoration with impulse noise removal, SIAM Journal on Imaging Sciences, 10 (2017), 1627-1667.  doi: 10.1137/16M1076034. [51] Y. Zhang, B. Dong and Z. Lu, An efficient algorithm for $\ell_{0}$ minimization in wavelet frame based image restoration, Mathematics of Computation, 82 (2013), 995-1015.  doi: 10.1090/S0025-5718-2012-02631-7. [52] W. Zhou, J. Cai and H. Gao, Adaptive tight frame based medical image reconstruction: A proof-of-concept study for computed tomography, Inverse Problems, 29 (2013), 125-146.  doi: 10.1088/0266-5611/29/12/125006. [53] L. Zhu, T. Niu and M. Petrongolo, Iterative CT reconstruction via minimizing adaptively reweighted total variation, Journal of X-ray Science and Technology, 22 (2014), 227-240.  doi: 10.3233/XST-140421.

Figures(11)

Tables(2)