Abstract:The aim of this article is to present a hybrid finite element/finite difference method which is used for reconstructions of electromagnetic properties within a realistic breast phantom. This is done by studying the mentioned properties' (electric permittivity and conductivity in this case) representing coefficients in a constellation of Maxwell's equations. This information is valuable since these coefficient can reveal types of tissues within the breast, and in applications could be used to detect shapes and locations of tumours.
Because of the ill-posed nature of this coefficient inverse problem, we approach it as an optimization problem by introducing the corresponding Tikhonov functional and in turn Lagrangian. These are then minimized by utilizing an interplay between finite element and finite difference methods for solutions of direct and adjoint problems, and thereafter by applying a conjugate gradient method to an adaptively refined mesh.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to reconstruct the complex permittivity function of the spatially distributed in a conductive medium on the boundary by using the time - varying electric field data of scattering. Specifically, this research aims to develop a hybrid finite element/finite - difference method (FE/FD) for reconstructing the electromagnetic properties in breast tissue, especially the dielectric constant and conductivity.
### Core of the problem
1. **Reconstructing electromagnetic properties**:
- By studying the coefficients (such as permittivity \(\varepsilon\) and conductivity \(\sigma\)) representing dielectric properties in Maxwell's equations to reveal different types of breast tissue.
- This information is of great value for detecting the shape and location of tumors.
2. **Challenges of the inverse problem**:
- Since this coefficient inverse problem is ill - posed, it needs to be treated as an optimization problem.
- Introduce the Tikhonov functional and the corresponding Lagrangian, and solve the forward problem and the adjoint problem by combining the finite element and finite - difference methods, and finally apply the conjugate gradient method for optimization.
3. **Practical applications**:
- Mainly applied to microwave medical imaging, especially breast cancer detection. By measuring the scattered electromagnetic waves, the dielectric properties of the internal structure of the breast can be reconstructed non - invasively.
### Formula expressions
The key formulas involved in the paper are as follows:
- **Maxwell's equations**:
\[
\mu \frac{\partial H}{\partial t}=-\nabla\times E \quad \text{in } \Omega_T,
\]
\[
\varepsilon \frac{\partial E}{\partial t}=\nabla\times H - \sigma E \quad \text{in } \Omega_T,
\]
\[
\nabla\cdot(\varepsilon E) = 0 \quad \text{in } \Omega_T.
\]
- **Tikhonov functional**:
\[
J(E,\varepsilon,\sigma):=\frac{1}{2}\|E - E_{\text{obs}}\|^2_{z,\Gamma_T}+\frac{\gamma_\varepsilon}{2}\|\varepsilon - \varepsilon_0\|^2_\Omega+\frac{\gamma_\sigma}{2}\|\sigma - \sigma_0\|^2_\Omega,
\]
where \(E_{\text{obs}}\) is the observed value on the boundary, \(\gamma_\varepsilon\) and \(\gamma_\sigma\) are regularization parameters, and \(\varepsilon_0\) and \(\sigma_0\) are initial guess values.
- **Adjoint problem**:
\[
\varepsilon \frac{\partial^2 \lambda}{\partial t^2}-\sigma \frac{\partial \lambda}{\partial t}-(\varepsilon^{-1})\nabla(\nabla\cdot\lambda)-\Delta\lambda = 0 \quad \text{in } \Omega_T,
\]
\[
\lambda(\cdot,T)=\frac{\partial \lambda}{\partial t}(\cdot,T) = 0 \quad \text{in } \Omega,
\]
\[
\frac{\partial \lambda}{\partial t}