Bayesian inversion for electromyography using low-rank tensor formats

DOI10.1088/1361-6420/ABD85AMaRDI QIDQ5859752zbMATHOpenAlexFDO

Authors Anna Rörich, Tim A. Werthmann, Dominik Göddeke, L. Grasedyck

Publication date 20 April 2021

Published in Inverse Problems (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2009.02772

zbMATH Keywords

inverse problemMetropolis-Hastings algorithmhierarchical Tucker formatEMGparameter-dependent problem

Mathematics Subject Classification ID

Bayesian inference (62F15) Monte Carlo methods (65C05) Computational methods for sparse matrices (65F50) Biomedical imaging and signal processing (92C55) Vector and tensor algebra, theory of invariants (15A72) Numerical methods for inverse problems for boundary value problems involving PDEs (65N21) Biological applications of optics and electromagnetic theory (78A70) Existence problems for PDEs: global existence, local existence, non-existence (35A01) Uniqueness problems for PDEs: global uniqueness, local uniqueness, non-uniqueness (35A02) Inverse problems (including inverse scattering) in optics and electromagnetic theory (78A46) Monte Carlo methods applied to problems in optics and electromagnetic theory (78M31)

Abstract: The reconstruction of the structure of biological tissue using electromyographic data is a non-invasive imaging method with diverse medical applications. Mathematically, this process is an inverse problem. Furthermore, electromyographic data are highly sensitive to changes in the electrical conductivity that describes the structure of the tissue. Modeling the inevitable measurement error as a stochastic quantity leads to a Bayesian approach. Solving the discretized Bayes-inverse problem means drawing samples from the posterior distribution of parameters, e.g., the conductivity, given measurement data. Using, e.g., a Metropolis-Hastings algorithm for this purpose involves solving the forward problem for different parameter combinations which requires a high computational effort. Low-rank tensor formats can reduce this effort by providing a data-sparse representation of all occurring linear systems of equations simultaneously and allow for their efficient solution. The application of Bayes' theorem proves the well-posedness of the Bayes-inverse problem. The derivation and proof of a low-rank representation of the forward problem allow for the precomputation of all solutions of this problem under certain assumptions, resulting in an efficient and theory-based sampling algorithm. Numerical experiments support the theoretical results, but also indicate that a high number of samples is needed to obtain reliable estimates for the parameters. The Metropolis-Hastings sampling algorithm, using the precomputed forward solution in a tensor format, draws this high number of samples and therefore enables solving problems which are infeasible using classical methods.

Recommendations

• Sampling-free Bayesian inversion with adaptive hierarchical tensor representations
• Bayesian inversion in biomedical imaging
• Stochastic collocation algorithms using \(l_1\)-minimization for Bayesian solution of inverse problems
• Accelerating the Bayesian inference of inverse problems by using data-driven compressive sensing method based on proper orthogonal decomposition
• Bayesian inversion for electrical-impedance tomography in medical imaging using the nonlinear Poisson-Boltzmann equation

Cites work

• A literature survey of low-rank tensor approximation techniques
• A new scheme for the tensor representation
• A sensitivity study of conductivity values in the passive bidomain equation
• Algorithm 941: \texttt{htucker} -- a Matlab toolbox for tensors in hierarchical Tucker format
• Analysis of individual differences in multidimensional scaling via an \(n\)-way generalization of ``Eckart-Young decomposition
• Approximate iterations for structured matrices
• Approximation and sampling of multivariate probability distributions in the tensor train decomposition
• Complexity analysis of accelerated MCMC methods for Bayesian inversion
• Distributed hierarchical SVD in the hierarchical Tucker format.
• Efficient estimation of cardiac conductivities via POD-DEIM model order reduction
• Hierarchical Singular Value Decomposition of Tensors
• Inverse problems: a Bayesian perspective
• Iterative methods based on soft thresholding of hierarchical tensors
• Low-rank tensor Krylov subspace methods for parametrized linear systems
• Markov Chains
• On low-rank approximability of solutions to high-dimensional operator equations and eigenvalue problems
• On the Inverse of the Sum of Matrices
• Quasi-Monte Carlo and multilevel Monte Carlo methods for computing posterior expectations in elliptic inverse problems
• Sampling-free Bayesian inversion with adaptive hierarchical tensor representations
• Source localization in electromyography using the inverse potential problem
• Tensor Rank and the Ill-Posedness of the Best Low-Rank Approximation Problem
• Tensor spaces and numerical tensor calculus
• Tensor-sparsity of solutions to high-dimensional elliptic partial differential equations

Cited in

(1)

• State estimation in nonlinear parametric time dependent systems using tensor train

Describes a project that uses

Uses Software

• htucker

This page was built for publication: Bayesian inversion for electromyography using low-rank tensor formats

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5859752)