Tomographic image reconstruction using training images

DOI10.1016/J.CAM.2016.09.019zbMATH Open1353.65017arXiv1503.01993OpenAlexW2963345358WikidataQ56334257 ScholiaQ56334257MaRDI QIDQ344271FDOQ344271

Authors: Sara Soltani, Martin S. Andersen, Per Christian Hansen 

Publication date: 22 November 2016

Published in: Journal of Computational and Applied Mathematics (Search for Journal in Brave)

Abstract: We describe and examine an algorithm for tomographic image reconstruction where prior knowledge about the solution is available in the form of training images. We first construct a nonnegative dictionary based on prototype elements from the training images; this problem is formulated as a regularized non-negative matrix factorization. Incorporating the dictionary as a prior in a convex reconstruction problem, we then find an approximate solution with a sparse representation in the dictionary. The dictionary is applied to non-overlapping patches of the image, which reduces the computational complexity compared to other algorithms. Computational experiments clarify the choice and interplay of the model parameters and the regularization parameters, and we show that in few-projection low-dose settings our algorithm is competitive with total variation regularization and tends to include more texture and more correct edges.

Full work available at URL: https://arxiv.org/abs/1503.01993

Recommendations

• A tensor-based dictionary learning approach to tomographic image reconstruction
• Joint Gaussian dictionary learning and tomographic reconstruction
• Sparse shape reconstruction
• Joint-MAP tomographic reconstruction with patch similarity based mixture prior model
• Deep Learning--Based Dictionary Learning and Tomographic Image Reconstruction

zbMATH Keywords

computational complexityregularizationalgorithminverse problemsimage reconstructionnumerical exampletomographysparse representationdictionary learning

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Ill-posedness and regularization problems in numerical linear algebra (65F22) Biomedical imaging and signal processing (92C55) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Numerical aspects of computer graphics, image analysis, and computational geometry (65D18)

Cites Work

• Templates for convex cone problems with applications to sparse signal recovery
• $rm K$-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation
• Distributed optimization and statistical learning via the alternating direction method of multipliers
• Non-negative matrix factorization with sparseness constraints
• Sparse and redundant representations. From theory to applications in signal and image processing.
• From Sparse Solutions of Systems of Equations to Sparse Modeling of Signals and Images
• Learning the parts of objects by non-negative matrix factorization
• AIR tools -- a MATLAB package of algebraic iterative reconstruction methods
• Edge-preserving and scale-dependent properties of total variation regularization
• Linear and nonlinear inverse problems with practical applications
• An alternating direction algorithm for matrix completion with nonnegative factors
• Online learning for matrix factorization and sparse coding
• Discrete inverse problems. Insight and algorithms.
• Augmented Lagrangian-based sparse representation method with dictionary updating for image deblurring
• Learning Multiscale Sparse Representations for Image and Video Restoration
• A tensor-based dictionary learning approach to tomographic image reconstruction
• Dictionary Learning Algorithms for Sparse Representation
• Multi-frame compression: theory and design.
• Implementation of an optimal first-order method for strongly convex total variation regularization
• Bayesian Nonparametric Dictionary Learning for Compressed Sensing MRI
• The Radon transform and medical imaging

Cited In (11)

• Sparse shape reconstruction
• Nonnegative tensor patch dictionary approaches for image compression and deblurring applications
• Joint Gaussian dictionary learning and tomographic reconstruction
• Joint-MAP tomographic reconstruction with patch similarity based mixture prior model
• Deep Learning--Based Dictionary Learning and Tomographic Image Reconstruction
• A tensor-based dictionary learning approach to tomographic image reconstruction
• An incomplete project data image reconstruction algorithm based on dictionary learning method
• Unsupervised knowledge-transfer for learned image reconstruction
• Template-based image reconstruction from sparse tomographic data
• Biomedical image reconstruction: from the foundations to deep neural networks
• Computed tomography reconstruction using deep image prior and learned reconstruction methods

Uses Software

• AIR tools
• TFOCS
• DLMRI-Lab
• PyHST2

This page was built for publication: Tomographic image reconstruction using training images

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q344271)