Fast absolute 3D CGO-based electrical impedance tomography on experimental tank data.

Area of Science:

• Electrical Impedance Tomography (EIT)
• Computational Imaging
• Biomedical Engineering

Background:

• Absolute EIT aims to reconstruct EIT images directly from measurements.
• Traditional methods often struggle with accuracy and robustness, especially with imperfect models.

Purpose of the Study:

• To present the first 3D absolute EIT reconstructions using the CGO-based method on experimental data.
• To compare CGO-based methods against traditional TV-regularized non-linear least squares (TV-NLS) methods.
• To assess the robustness of CGO-based methods to domain shape modeling errors.

Main Methods:

• Development and application of 3D CGO-based algorithms for absolute EIT reconstruction.
• Comparison with TV-regularized non-linear least squares reconstruction.
• Robustness testing using incorrect domain shape models.

Main Results:

• CGO-based methods achieved fast (1-5 s) 3D absolute EIT reconstructions.
• Demonstrated strong robustness to incorrect domain modeling, comparable to difference EIT.
• Exhibited fewer boundary artifacts than TV-regularized NLS reconstructions.

Conclusions:

• This study presents the first experimental validation of 3D CGO-based absolute EIT.
• CGO-based methods show significant advantages in speed, robustness, and artifact reduction.
• The results encourage further development of CGO-based methods for practical absolute EIT applications.