Advanced Methods for High Resolution Electrical Impedance Tomography (EIT)

NU 2018-201

INVENTORS

• Matthew Grayson*
• Can Aygen
• Charles Costakis
• Lauren Lang
• Claire Cecelia Onsager
• John Rogers
• Andreas Tzavelis
• Chulin Wang

SHORT DESCRIPTION
Advanced methods for electrical impedance tomography (EIT) and resistance-based sensing, offering significant improvements in image resolution, noise tolerance, and computational efficiency.

BACKGROUND
Traditional tomographic methods for mapping resistivity, such as Electrical Resistance Tomography (ERT) and Electrical Impedance Tomography (EIT), have long been used for geological and medical applications. Traditional approaches often rely on computationally intensive iterative solvers and finite-element models that can yield ill-conditioned inverse problems, particularly when dealing with noisy data or intricate electrode configurations. Additionally, standard methods frequently struggle with the challenge of mapping sparse, discrete measurements to continuous, high-fidelity images. Thus, traditional methods have robustness and scalability problems when applied to diverse applications ranging from biomedical imaging to structural monitoring and wearable sensing.

ABSTRACT
The novel approach of this invention utilizes optimized electrical impedance tomography techniques for both 2D and 3D applications by representing spatial resistivity variations as a weighted sum of orthogonal basis functions—typically Zernike polynomials for circular surfaces and spherical harmonics for spherical volumes—with the option to refine these functions using principal component analysis or other constraints. It enhances measurement fidelity by optimizing electrode placement and selecting a reduced yet effective subset of independent tetra-polar resistance measurements to form a well-conditioned sensitivity matrix, whose optimization is based on maximizing the corresponding Jacobian determinant. The reconstruction process addresses the inverse problem through iterative refinement methods such as truncated singular value decomposition or regularized inverse solutions while maintaining computational efficiency by aligning the number of basis functions with the available independent measurements.

APPLICATIONS

• Biomedical Imaging and Wearables
  • Provides quantitative metrics to fine-tune electrode placement and higher resolution measurement strategies, leading to enhanced performance in medical applications.
• Translational Technology
  • Advanced EIT methods can be similarly optimized for industrial applications, such as artificial touch-sensitive skin for robotics and geological surveying.

ADVANTAGES

• High Resolution:  achieves a significantly higher grid resolution compared to existing methods.
• Optimized System Design: qualitatively informs electrode placement and measurement selection.
• Computationally Efficient:  utilizes minimal orthogonal basis functions for rapid and efficient mapping.
• Adaptability: efficiently evaluates electrode configurations to provide optimized electrode placement measurement strategies for non-standard applications/configurations.

IP STATUS

US Patent Application #17/295,318

US Patent Application #18/049,551