Study on the accuracy of forward modeling in electrical impedance tomography for thorax imaging

In this paper, we study the data calibration procedure in electrical impedance tomography (EIT) for thorax imaging. Data calibration is an important procedure in EIT imaging as it bridges the gap between simulated data and measured data. Minimizing the systematic error caused by forward modeling could significantly improve the quality of reconstruction. A few factors are investigated in this study including: 1) electrode model; 2) skin impedance; 3) dimensionality of forward modeling. We observe that the dimensionality of forward modeling is the most critical factor. In fact, simulated data comparable with the experimental data can be obtained using three-dimensional forward modeling and reasonable estimation of background conductivity. It can serve as a good starting point for EIT data inversion. Finally, we perform a fully three-dimensional reconstruction using the measured data of human pulmonary ventilation. The reconstructed images further support our conclusion.