Statistical inversion in electrical impedance tomography using mixed total variation and non-convex lp regularization prior

Thilo Strauss; Taufiquar Khan

doi:10.1515/jiip-2013-0064

Statistical inversion in electrical impedance tomography using mixed total variation and non-convex l_p regularization prior

Thilo Strauss, Taufiquar Khan

Research output: Contribution to journal › Article › peer-review

16 Citations (Scopus)

Abstract

Electrical impedance tomography (EIT) is a well-known technique to estimate the conductivity distribution γ of a body Ω with unknown electromagnetic properties. EIT is a severely ill-posed inverse problem. In this paper, we formulate the EIT problem in the Bayesian framework using mixed total variation (TV) and non-convex l_p regularization prior. We use the Markov Chain Monte Carlo (MCMC) method for sampling the posterior distribution to solve the ill-posed inverse problem in EIT. We present simulations to estimate the distribution for each pixel for the image reconstruction of the conductivity in EIT.

Original language	English
Pages (from-to)	529-542
Number of pages	14
Journal	Journal of Inverse and Ill-Posed Problems
Volume	23
Issue number	5
DOIs	https://doi.org/10.1515/jiip-2013-0064
Publication status	Published - 1 Oct 2015
Externally published	Yes

Keywords

EIT
electrical impedance tomography
inverse problem
Markov Chain Monte Carlo method (MCMC)
Metropolis-Hastings algorithm
mixed TV and l prior
non-convex l regularization
non-convex prior
statistical inversion
total variation (TV) prior

Access to Document

10.1515/jiip-2013-0064

Cite this

@article{40daac03a87043bfa16a9420a02f42a4,

title = "Statistical inversion in electrical impedance tomography using mixed total variation and non-convex lp regularization prior",

abstract = "Electrical impedance tomography (EIT) is a well-known technique to estimate the conductivity distribution γ of a body Ω with unknown electromagnetic properties. EIT is a severely ill-posed inverse problem. In this paper, we formulate the EIT problem in the Bayesian framework using mixed total variation (TV) and non-convex lp regularization prior. We use the Markov Chain Monte Carlo (MCMC) method for sampling the posterior distribution to solve the ill-posed inverse problem in EIT. We present simulations to estimate the distribution for each pixel for the image reconstruction of the conductivity in EIT.",

keywords = "EIT, electrical impedance tomography, inverse problem, Markov Chain Monte Carlo method (MCMC), Metropolis-Hastings algorithm, mixed TV and l prior, non-convex l regularization, non-convex prior, statistical inversion, total variation (TV) prior",

author = "Thilo Strauss and Taufiquar Khan",

note = "Publisher Copyright: {\textcopyright} 2015 by De Gruyter.",

year = "2015",

month = oct,

day = "1",

doi = "10.1515/jiip-2013-0064",

language = "English",

volume = "23",

pages = "529--542",

journal = "Journal of Inverse and Ill-Posed Problems",

issn = "0928-0219",

number = "5",

}

TY - JOUR

T1 - Statistical inversion in electrical impedance tomography using mixed total variation and non-convex lp regularization prior

AU - Strauss, Thilo

AU - Khan, Taufiquar

PY - 2015/10/1

Y1 - 2015/10/1

N2 - Electrical impedance tomography (EIT) is a well-known technique to estimate the conductivity distribution γ of a body Ω with unknown electromagnetic properties. EIT is a severely ill-posed inverse problem. In this paper, we formulate the EIT problem in the Bayesian framework using mixed total variation (TV) and non-convex lp regularization prior. We use the Markov Chain Monte Carlo (MCMC) method for sampling the posterior distribution to solve the ill-posed inverse problem in EIT. We present simulations to estimate the distribution for each pixel for the image reconstruction of the conductivity in EIT.

AB - Electrical impedance tomography (EIT) is a well-known technique to estimate the conductivity distribution γ of a body Ω with unknown electromagnetic properties. EIT is a severely ill-posed inverse problem. In this paper, we formulate the EIT problem in the Bayesian framework using mixed total variation (TV) and non-convex lp regularization prior. We use the Markov Chain Monte Carlo (MCMC) method for sampling the posterior distribution to solve the ill-posed inverse problem in EIT. We present simulations to estimate the distribution for each pixel for the image reconstruction of the conductivity in EIT.

KW - EIT

KW - electrical impedance tomography

KW - inverse problem

KW - Markov Chain Monte Carlo method (MCMC)

KW - Metropolis-Hastings algorithm

KW - mixed TV and l prior

KW - non-convex l regularization

KW - non-convex prior

KW - statistical inversion

KW - total variation (TV) prior

UR - http://www.scopus.com/inward/record.url?scp=84943631694&partnerID=8YFLogxK

U2 - 10.1515/jiip-2013-0064

DO - 10.1515/jiip-2013-0064

M3 - Article

AN - SCOPUS:84943631694

SN - 0928-0219

VL - 23

SP - 529

EP - 542

JO - Journal of Inverse and Ill-Posed Problems

JF - Journal of Inverse and Ill-Posed Problems

IS - 5

ER -

Statistical inversion in electrical impedance tomography using mixed total variation and non-convex l_p regularization prior

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this