A Comparative Study on Performances of Adaptive and Nonadaptive Sparse Solvers for Electrical Impedance Tomography

In electrical impedance tomography (EIT), one reconstructs the electric impedance of a part of an object or human body from its surface measurements of voltages. Applications of EIT are far and wide in several fields, including medical imaging, nondestructive testing and process tomography. Reconstruction in EIT in practical settings usually deals with a finite number of voltage differences, which is one of the standard problems in EIT. Of late, sparsity-based optimization techniques have been shown to provide economic descriptions in the sense that reconstruction of the underlying signal (or image) can be made from a few of its linear measurements. The adaptive sparse solvers are proven to be even more effective. Driven by the recent surge in nonconvex minimization, we deploy several solvers in the reconstruction of EIT and then compare and contrast the significance of adaptive minimization against standard solvers in EIT reconstruction.