Magnetic resonance electrical properties tomography for small anomalies using boundary conditions: A simulation study

Joonsung Lee, Narae Choi, Jin Keun Seo, Dong Hyun Kim

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)


Purpose: Magnetic resonance electrical property tomography (MREPT) is an emerging imaging modality using measured B1 maps from magnetic resonance imaging (MRI) to measure a distribution of electric conductivity and permittivity of the subject at the Larmor frequency. Conventional MREPT approaches at single transmit channel system using the Helmholtz equation rely on an assumption that conductivity and permittivity of the subject are locally homogeneous. For small tissue structures and tissue boundaries, in which the assumption of locally homogeneous conductivity and permittivity does not hold, the reconstructed conductivity values deviated from the actual values, so called "Boundary Artifacts." The aim of this study is to propose new reconstruction processes based on time-harmonic Maxwell's equations to reconstruct conductivity for small tissue structures and tissue boundaries. Methods: Instead of removing the electric fields from the equations as done in the Helmholtz equation, three key identities of circularly polarized and longitudinal components of electric fields, circularly polarized component of magnetic fields, and electric properties from time-harmonic Maxwell's equations are derived. Based on the three key identities, the proposed reconstruction methods determine conductivity, permittivity, and circularly polarized component and longitudinal component of electric fields using the measured H1+. In each iterative step, estimated conductivity, permittivity, electric fields, and artifact-free mask region, Ω, where the contribution of the boundary artifacts is small, were updated. Using the estimated values in the artifact-free mask region as boundary conditions, the estimates beyond the mask region were updated. EM simulations were performed on three types of numerical phantoms with very small regions of homogeneous conductivity and permittivity. The performance of the proposed methods was evaluated using the simulated electric and magnetic fields. Results: For the numerical simulation model, the proposed methods significantly reduced the boundary artifacts compared to conventional methods using Helmholtz equations. In addition, previous methods using the Helmholtz equation could measure conductivity of only large anomalies, but the proposed method can measure the conductivity of the small compartments whose size is 2-3 voxels. The proposed approaches are compatible with spatial filtering which can be used to reduce noise. If a good image segmentation is available as a prior information, better initial boundary conditions can be estimated, and thus the proposed approach can be more accurate for small tissue structures. Conclusions: The proposed reconstruction method not only determines electrical properties, but also circularly polarized component and longitudinal component of electric fields using an iterative process. The proposed method can quantitatively detect the conductivity of the small anomalies better than conventional methods.

Original languageEnglish
Pages (from-to)4773-4785
Number of pages13
JournalMedical physics
Issue number9
Publication statusPublished - 2017 Sept

Bibliographical note

Funding Information:
This work was supported by IBS-R015-D1 for J. Lee and Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and future Planning (NRF-2016R1 A2B3016273 for N. Choi, and D.-H. Kim, and NRF-2015R 1A5A1009350 for J. K. Seo).

Publisher Copyright:
© 2017 American Association of Physicists in Medicine.

All Science Journal Classification (ASJC) codes

  • Biophysics
  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Magnetic resonance electrical properties tomography for small anomalies using boundary conditions: A simulation study'. Together they form a unique fingerprint.

Cite this