Reconstruction of electrical impedance tomography images using chaotic self-adaptive ring-topology differential evolution and genetic algorithms

The exposition of living tissues to ionizing radiation can result on several health problems, increasing the probability of cancer. Efforts from both academy and industry to develop and improve non-invasive methods have been increasing since the 1990´s. Electrical Impedance Tomography (EIT) is a noninvasive imaging technique that offers a vast field of possibilities for imaging diagnostics, once it is a low cost, portable, and safe of handling technology. Nevertheless, EIT image reconstruction is an ill-posed problem: there are no unique mathematical solutions to solve the Equation of Poison. Herein this work we present an EIT reconstruction method based on the finite-element method and the optimization of the relative error of reconstruction using Self-Adaptive Ring-Topology Differential Evolution (SRDE) and its modified version using chaotic mutation factor (CSRDE). Our proposal was compared with genetic algorithms and classical differential evolution strategies, considering initial populations of 100 individuals. CSRDE-based experiments were ran using 70 agents evolving by SRDE and 30 chaotic mutated agents generated from the 30 worst agents. Results were quantitatively evaluated with ground-truth images using the relative mean squared error, demonstrating that our results using CSRDE reached considerably low error magnitudes. Qualitative evaluation also indicated that our results were anatomically consistent.