Current Density Impedance Imaging and Least Gradient Problems

Current density impedance imaging (CDII) is a hybrid modality that merges applied electrical currents with magnetic resonance measurements to map the internal conductivity distribution of a medium. By injecting low‐amplitude currents and using magnetic resonance to measure the resulting current density field, one obtains interior data that greatly enhances the stability of the inverse conductivity problem. Reconstruction then reduces to solving a partial differential equation of the form div(σ∇u)=0, where σ is the sought conductivity and u the electric potential, coupled with measured values of the current density J=−σ∇u. This leads naturally to a variational formulation in which one seeks a potential whose gradient minimises a weighted measure of variation, subject to boundary and interior data constraints. Such formulations are known as least gradient problems: among all functions matching prescribed boundary values, one selects that which minimises the total variation or an anisotropic generalisation. These problems pose challenging questions of existence, uniqueness, stability and regularity, especially when data are only partially available or when anisotropic conductivities are involved. Advances in this field have broad impact, ranging from improved tumour detection in clinical MRI to enhanced subsurface imaging in geophysics, and underpin developments in nondestructive testing and material science.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Recent studies have addressed the mixed Dirichlet–Neumann least gradient problem in the plane by establishing an equivalence with optimal transport formulations. This has enabled proofs of existence, uniqueness and Sobolev‐type regularity for solutions when part of the boundary carries prescribed fluxes and the remainder fixed potentials. In a metric‐measure‐space setting, researchers have extended the classical Dirichlet problem by allowing boundary data in L1 and by developing nonlocal techniques to accommodate discontinuities and irregular domains. They have shown that solutions exist for data approximable by continuous functions, and that most boundary values admit a matching least gradient interior solution. Foundational work on planar domains with convex and nonconvex geometries has clarified how partial boundary conditions determine solution structure; these analyses reveal that strict convexity guarantees uniqueness and stability, while more complex domains may admit multiple minimisers or facet detachments at concave boundaries.

Current Density Impedance Imaging and Least Gradient Problems publication trend

The graph below shows the total number of articles in current density impedance imaging and least gradient problems across all publications each year (not limited to Nature Index journals).

Technical terms

Electrical conductivity (σ): Measure of a medium’s ability to conduct electric current.

Current density (J): Vector field representing electric current per unit area, given by J=−σ∇u in conductive media.

Inverse conductivity problem: The task of recovering σ from boundary or interior observations of u or J.

Least gradient problem: Variational problem seeking a function that minimises the L1 norm of its gradient subject to prescribed boundary or interior data.

Total variation: Integral of the magnitude of the gradient; in imaging, it promotes piecewise-constant reconstructions.

Dirichlet boundary condition: Constraint prescribing the value of the potential u on part of the boundary.

Neumann boundary condition: Constraint prescribing the normal derivative (flux) of u on part of the boundary.

Bounded variation (BV) space: Function space consisting of functions whose distributional derivatives are finite Radon measures, suitable for least gradient formulations.

References

1. Special cases of the planar least gradient problem. Nonlinear Analysis (2017).
2. The least gradient problem with Dirichlet and Neumann boundary conditions. Advances in Calculus of Variations (2024).
3. Non-locality, non-linearity, and existence of solutions to the Dirichlet problem for least gradient functions in metric measure spaces. Revista Matemática Iberoamericana (2022).
4. A function space framework for structural total variation regularization with applications in inverse problems. Inverse Problems (2018).