Comparisons among EIT data collection techniques and reconstruction algorithms

Electrical Impedance Tomography (EIT) is an imaging technique that aims to reconstruct the spatial electrical conductivity distribution in sections of the human body. In this paper, in order to solve the EIT forward and inverse problems, a finite difference approach to the solution of Maxwell’s equations and the Newton-Raphson algorithm have been employed, respectively. In particular, the inverse problem has been solved using the Tikhonov regularization with various choices of the regularization matrix. Moreover, different data collection methods have been tested on simulated measurements. The obtained results have been compared based on the average deviation of the estimated conductivity distribution with respect to the reference one. The reconstruction procedure has been validated through a comparison with the EIDORS open source software. The best image reconstruction has been obtained by using the neighboring data collection method with null regularization matrix, and using the truncated singular value decomposition to perform the matrix inversion. Moreover, the cross and opposite data collection methods showed better performance than the neighboring one in the presence of a random noise added to the measured signal, while the opposite method evidenced the best results with respect to electrode positioning uncertainties.