论文信息 - Nonconvex optimization for improved exploitation of gradient sparsity in CT image reconstruction

Nonconvex optimization for improved exploitation of gradient sparsity in CT image reconstruction

A nonconvex optimization algorithm is developed, which exploits gradient magnitude image (GMI) sparsity for reduction in the projection view angle sampling rate. The algorithm shows greater potential for exploiting GMI sparsity than can be obtained by convex total variation (TV) based optimization. The nonconvex algorithm is demonstrated in simulation with ideal, noiseless data for a 2D fan-beam computed tomography (CT) configuration, and with noisy data for a 3D circular cone-beam CT configuration.

[1] Rick Chartrand,et al. Nonconvex Splitting for Regularized Low-Rank + Sparse Decomposition , 2012, IEEE Transactions on Signal Processing.

[2] Rick Chartrand,et al. Exact Reconstruction of Sparse Signals via Nonconvex Minimization , 2007, IEEE Signal Processing Letters.

[3] E.J. Candes,et al. An Introduction To Compressive Sampling , 2008, IEEE Signal Processing Magazine.

[4] Xiaochuan Pan,et al. Enhanced imaging of microcalcifications in digital breast tomosynthesis through improved image-reconstruction algorithms. , 2009, Medical physics.

[5] Xiaochuan Pan,et al. First-order convex feasibility algorithms for x-ray CT. , 2012, Medical physics.

[6] P. L. Combettes. The foundations of set theoretic estimation , 1993 .

[7] Xiaochuan Pan,et al. Image reconstruction from few views by non-convex optimization , 2007, 2007 IEEE Nuclear Science Symposium Conference Record.

[8] Antonin Chambolle,et al. A First-Order Primal-Dual Algorithm for Convex Problems with Applications to Imaging , 2011, Journal of Mathematical Imaging and Vision.

[9] Xiao Han,et al. Optimization-based reconstruction of sparse images from few-view projections , 2012, Physics in medicine and biology.