Compensation for geometric modeling errors by electrode movement in electrical impedance tomography

TL;DR

This paper demonstrates that electrode movement can compensate for geometric modeling errors in electrical impedance tomography, improving reconstruction accuracy in 2D by estimating electrode positions and sizes, supported by numerical and experimental evidence.

Contribution

It introduces a method to account for boundary uncertainties in EIT by estimating electrode parameters during reconstruction, with theoretical and numerical validation.

Findings

Electrode movement improves 2D reconstruction accuracy.

The complete electrode model is approximately conformally invariant.

The approach is less effective in 3D reconstructions.

Abstract

Electrical impedance tomography aims at reconstructing the conductivity inside a physical body from boundary measurements of current and voltage at a finite number of contact electrodes. In many practical applications, the shape of the imaged object is subject to considerable uncertainties that render reconstructing the internal conductivity impossible if they are not taken into account. This work numerically demonstrates that one can compensate for inaccurate modeling of the object boundary in two spatial dimensions by estimating the locations and sizes of the electrodes as a part of a reconstruction algorithm. The numerical studies, which are based on both simulated and experimental data, are complemented by proving that the employed complete electrode model is approximately conformally invariant, which suggests that the obtained reconstructions in mismodeled domains reflect conformal images of the true targets. The numerical experiments also confirm that a similar approach does not, in general, lead to a functional algorithm in three dimensions.