Identifying conductivity in electrical impedance tomography with total variation regularization (Q2413471)

The authors discuss a variational method with total variation regularization for identifying the conductivity in electrical impedance tomography from one boundary measurement. For solving the corresponding minimization problem numerically, the projected Armijo algorithm [\textit{C. T. Kelley}, Iterative methods for optimization. Philadelphia, PA: Society for Industrial and Applied Mathematics (1999; Zbl 0934.90082)]) is applied. In each iteration step of this algorithm one has to solve a Dirichlet and a Neumann boundary value problem. For the discretization of these problems the finite element method with continuous piecewise linear functions is used. The stability of the finite element method for the regularized identification problem is shown and the convergence of the regularized finite element approximations to a solution of the identification problem is proved. Three numerical examples are presented for illustrating the theoretical results.