Data-Driven 3D Reconstruction for Electrical Impedance Tomography

The use of data-driven reconstruction methods to solve inverse problems has steadily increased, particularly in the field of electrical impedance tomography (EIT). This trend is due to the advantages of machine learning (ML) in addressing ill-posed, non-linear inverse problems, like EIT image reconstruction. Recently, there has been an increase in research on two-dimensional EIT image reconstruction. However, the evaluation of data-driven three-dimensional image reconstruction has not been extensively examined. This contribution demonstrates the feasibility, reliability, and challenges of three-dimensional EIT imaging using experimental data.