Deep Neural Networks for Inverse Problems with Pseudodifferential Operators: An Application to Limited-Angle Tomography

DOI10.1137/20M1343075zbMath1489.44003arXiv2006.01620OpenAlexW3157843833MaRDI QIDQ5860291

Tatiana A. Bubba, Samuli Siltanen, Mathilde Galinier, Luca Ratti, Marco Prato, Matti Lassas

Publication date: 19 November 2021

Published in: SIAM Journal on Imaging Sciences (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2006.01620

zbMATH Keywords

waveletspseudodifferential operatorsFourier integral operatorsmicrolocal analysisX-ray transformsparse regularizationconvolutional neural networksdeep neural networkslimited-angle tomography

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Biomedical imaging and signal processing (92C55) Radon transform (44A12) Pseudodifferential and Fourier integral operators on manifolds (58J40) Fourier integral operators applied to PDEs (35S30)

Related Items (7)

Deep microlocal reconstruction for limited-angle tomography ⋮ Inverse problems on low-dimensional manifolds ⋮ Convergence Rates for Learning Linear Operators from Noisy Data ⋮ Convergence Results for Primal-Dual Algorithms in the Presence of Adjoint Mismatch ⋮ On Learning the Invisible in Photoacoustic Tomography with Flat Directionally Sensitive Detector ⋮ Tomographic inverse problems: mathematical challenges and novel applications. Abstracts from the workshop held April 30 -- May 5, 2023 ⋮ Solving inverse problems with deep learning

Uses Software

• scikit-image
• ASTRA
• Adam
• ODL
• GitHub
• PyWavelets
• tfwavelets
• ISTA-Net
• DeepAdverserialRegulariser

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Injectivity and \(\text{weak}^\star\)-to-weak continuity suffice for convergence rates in \(\ell^{1}\)-regularization
• Linear convergence of iterative soft-thresholding
• Nonlocal inversion formulas for the X-ray transform
• Inversion of incomplete Radon transform
• Wavelet approximation methods for pseudodifferential equations. I: Stability and convergence
• Local tomography for the limited-angle problem
• From error bounds to the complexity of first-order descent methods for convex functions
• On the strength of streak artifacts in filtered back-projection reconstructions for limited angle weighted X-ray transform
• The inverse problem for the local geodesic ray transform
• Estimates for singular Radon transforms and pseudodifferential operators with singular symbols
• The Mathematics of Computerized Tomography
• Mathematical Methods in Image Reconstruction
• Microlocal Analysis in Tomography
• The Ill-Conditioned Nature of the Limited Angle Tomography Problem
• Singularities of the X-Ray Transform and Limited Data Tomography in $\mathbb{R}^2 $ and $\mathbb{R}^3 $
• Fast wavelet transforms and numerical algorithms I
• Local Tomography for the Generalized Radon Transform
• Solving ill-posed inverse problems using iterative deep neural networks
• Deep Convolutional Neural Network for Inverse Problems in Imaging
• Directional sinogram inpainting for limited angle tomography
• Linear and Nonlinear Inverse Problems with Practical Applications
• An iterative thresholding algorithm for linear inverse problems with a sparsity constraint
• Solving inverse problems using data-driven models
• Analyzing Reconstruction Artifacts from Arbitrary Incomplete X-ray CT Data
• How strong are streak artifacts in limited angle computed tomography?
• Learning representations by back-propagating errors
• Learning the invisible: a hybrid deep learning-shearlet framework for limited angle computed tomography
• Characterization and reduction of artifacts in limited angle tomography
• Electron tomography: a short overview with an emphasis on the absorption potential model for the forward problem

This page was built for publication: Deep Neural Networks for Inverse Problems with Pseudodifferential Operators: An Application to Limited-Angle Tomography