Sparse reconstruction of log-conductivity in current density impedance tomography (Q2303268)

The current density impedance imaging (CDII), combines the classical electrical impedance tomography (EIT) setup with a magnetic resonance (MR) scanning. The current is applied through the electrodes which give rise to an interior electric field and the corresponding generated magnetic field is measured by the MR scanner. One considers the following conductivity equation \[ -\nabla.(e^{\sigma(x,y)}\nabla u(x,y))=0\text{ in }\Omega,\quad u(x,y)|_{\Gamma} = f_D(x,y) \] where $\Gamma\subset \mathbb{R}^2$, $\Gamma$ is the boundary of $\Omega$, $e^{\sigma}$ is the conductivity coefficient and $u$ is the electrical potential. The problem is to recover the conductivity $\sigma$ from the current $f$ applied to the boundary. This way, the CDII inversion problem is viewed as an inverse problem of estimating the conductitity from the above PDE. The problem is solved by an optimization approach. A cost-functional is introduced, one formulates a minimization problem, existence results and first-order necessary conditions are proved. Results on numerical implementation are largely discussed.