🔢 mathematics

Infinite-dimensional Lipschitz stability in the Calderón problem and general Zernike bases

This paper establishes that the Calderón problem admits infinite-dimensional Lipschitz stability on the unit ball by proving that conductivities with rapidly decaying coefficients in general $d$ -dimensional Zernike bases—allowing full angular freedom but restricted radial variation—can be uniquely and stably recovered.

Original authors: Henrik Garde, Markus Hirvensalo, Nuutti Hyvönen

Published 2026-02-18

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Henrik Garde, Markus Hirvensalo, Nuutti Hyvönen

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: The "Black Box" Mystery

Imagine you have a mysterious, opaque ball (like a black marble). You cannot see inside it. However, you can touch the surface and perform a specific experiment:

The Input: You apply a voltage (electric potential) to different spots on the surface.
The Output: You measure the resulting electrical current flowing out of the surface.

The Goal: Based only on these surface measurements, can you figure out exactly what the material is made of inside the ball? Is it copper? Is it plastic? Is it a weird mix of both?

This is the Calderón Problem (or the Inverse Conductivity Problem). It's like trying to guess the ingredients of a cake just by tasting the frosting, without ever cutting a slice.

The Problem: It's Usually Impossible to Be Precise

For a long time, mathematicians knew that for a general ball with any kind of material inside, this problem is incredibly unstable.

The Analogy: Imagine trying to guess the exact recipe of a soup by tasting a single spoonful. If you change the amount of salt by a tiny, invisible amount, the taste might change drastically. Or, conversely, two completely different soups might taste almost identical.
The Math: In the worst-case scenario, a tiny error in your measurement leads to a massive, wild guess about the inside. This is called Logarithmic Stability. It means the problem is "ill-posed" and very hard to solve reliably.

The Breakthrough: Finding a "Safe Zone"

The authors of this paper asked a clever question: "What if we don't try to guess ANY possible material? What if we restrict our guess to a specific family of materials?"

They discovered that there are infinite families of materials (infinite-dimensional classes) where the problem becomes stable and solvable.

The Analogy: Instead of guessing any soup in the world, you agree that the soup is always a "Tomato Basil" soup, but you just don't know the exact ratio of tomatoes to basil. If you know the recipe structure is fixed, a tiny taste change tells you exactly how much salt was added. You can now predict the inside with high precision. This is called Lipschitz Stability.

The Secret Weapon: The "Zernike" Recipe Book

How did they find these stable families? They used a special mathematical tool called Zernike Bases.

Think of the inside of the ball as a complex 3D painting. To describe this painting, you need a "vocabulary."

Old Vocabulary: You could try to describe it by listing the color of every single pixel. That's too much data and too messy.
The Zernike Vocabulary: This is like a set of standard "brushstrokes" or "patterns" (like swirls, rings, and waves) that you can mix together to build any shape.
- Some patterns describe the direction (like the swirl of a galaxy).
- Some patterns describe the distance from the center (like the rings of a tree).

The authors proved that if you build your "mystery material" using these specific Zernike brushstrokes, you can reconstruct the inside perfectly from the outside measurements.

The Catch: The "Edge" Rule

There is one rule for this to work. The authors found that the "brushstrokes" (the mathematical coefficients) must fade away quickly as you get closer to the edge of the ball.

The Analogy: Imagine the material inside the ball is a sound. If the sound gets incredibly loud right at the very edge of the container, it creates a chaotic echo that messes up your measurements. But, if the sound fades out gently as it approaches the edge, the echo is clean, and you can hear exactly what the sound was.
The Math: The paper proves that if the "volume" of these Zernike patterns drops off fast enough near the boundary, the math works out, and the stability is guaranteed.

Why Does This Matter?

Medical Imaging: This is crucial for things like Electrical Impedance Tomography (EIT), used to image lungs or brains. If we know the tissue structure follows certain patterns (like the Zernike ones), doctors can get much sharper, more reliable images without needing dangerous radiation (like X-rays).
Geology: It helps in figuring out what's underground (oil, water, rock) by measuring electrical signals on the surface.
Mathematical History: Before this, people thought you could only get this kind of stability if you limited yourself to a finite number of patterns (like only using 5 brushstrokes). This paper proves you can use infinite patterns and still get a perfect, stable result.

Summary in One Sentence

The authors found a special "dictionary" (Zernike bases) for describing materials inside a sphere, proving that if you use this dictionary and ensure the patterns fade out nicely at the edge, you can perfectly reconstruct the hidden interior from surface measurements, turning a chaotic guessing game into a precise science.

1. Problem Statement

The Calderón inverse conductivity problem asks whether the electrical conductivity coefficient $\gamma$ inside a domain $B$ can be uniquely determined from boundary measurements given by the Dirichlet-to-Neumann (DtN) map $\Lambda(\gamma)$ .

Stability Context: For general infinite-dimensional classes of conductivities, the problem is known to be conditionally logarithmically stable (the error in $\gamma$ decays logarithmically with respect to the error in $\Lambda$ ). This is considered optimal for general settings but implies severe ill-posedness.
Finite-Dimensional Case: Lipschitz stability (linear error decay) is well-established for finite-dimensional classes of conductivities.
The Gap: Prior to this work, it was an open question whether Lipschitz stability could be achieved for infinite-dimensional classes of conductivities in the nonlinear Calderón problem. Most existing infinite-dimensional results were limited to linearized settings or specific inclusion detection problems.

2. Methodology

The authors develop a framework to identify specific infinite-dimensional subspaces of conductivities where Lipschitz stability holds. The methodology relies on three main pillars:

A. Boundary Determination and Maximum Principles

The authors leverage a classical result by Sylvester and Uhlmann (Theorem 1.1), which establishes Lipschitz stability for determining the conductivity on the boundary ( $\partial B$ ).

Key Insight: If the difference between two conductivities, $\delta = \gamma_1 - \gamma_2$ , satisfies a weak maximum principle (e.g., if $\delta$ is a solution to an elliptic equation with a non-negative zero-order term), then the $L^\infty$ norm of $\delta$ in the interior is bounded by its norm on the boundary.
Strategy: Construct infinite-dimensional classes of conductivities where the difference $\gamma_1 - \gamma_2$ is constrained such that its interior behavior is controlled by its boundary trace.

B. General Basis Expansion with Decay Conditions

The authors propose expanding the conductivity difference in a general orthonormal basis $\{\psi_{n,k}\}$ defined on the unit ball $B \subset \mathbb{R}^d$ :
$\psi_{n,k}(r\theta) = g_{n,k}(r) f_n(\theta)$
where $f_n$ are spherical harmonics (angular part) and $g_{n,k}$ are radial functions.

Assumptions: The basis functions must satisfy specific regularity and boundary behavior conditions (Assumption 1.4), particularly that $|g_{n,k}(1)|$ is bounded away from zero.
Weighted Decay: To ensure the boundary trace of the infinite series exists, the expansion coefficients $(c_{n,k})$ must belong to a weighted $\ell^1$ space.
The Stability Condition: The core technical contribution is Condition (1.2), which restricts the "freedom" of the coefficients in the radial direction. It requires that for any fixed angular frequency $n$ , the coefficients $c_{n,k}$ do not cancel each other out excessively. Specifically:
$\sum_k |c_{n,k}|^2 \leq \frac{1}{\epsilon^2} \left| \sum_k c_{n,k} g_{n,k}(1) \right|^2$
This condition implies that the radial variation of the conductivity is limited, preventing the "smoothing out" of boundary information that typically leads to logarithmic instability.

C. Construction of General Zernike Bases

To provide concrete examples satisfying the above assumptions, the authors construct $d$ -dimensional Zernike bases for $d \geq 2$ .

Derivation: They define a Zernike operator $Z = -\Delta + (x \cdot \nabla)^2 + d(x \cdot \nabla)$ and solve the associated eigenvalue problem using separation of variables.
Radial Part: The radial eigenfunctions are shown to be related to Gauss hypergeometric functions and can be expressed as Jacobi polynomials (specifically, radial Zernike polynomials).
Verification: They prove that these constructed bases satisfy all conditions of Assumption 1.4, including the crucial lower bound on the boundary values $|g_{n,k}(1)|$ .

3. Key Contributions

First Infinite-Dimensional Lipschitz Stability Result: The paper provides the first proof that Lipschitz stability holds for infinite-dimensional classes of conductivities in the nonlinear Calderón problem.
General Stability Theorem (Theorem 1.6): A general theorem stating that if the conductivity difference belongs to a specific subspace $A$ (defined by the basis expansion and decay condition 1.2), then:
$\|\gamma_1 - \gamma_2\|_{L^2(B)} \leq C \|\Lambda(\gamma_1) - \Lambda(\gamma_2)\|_{\mathcal{L}(H^{1/2}, H^{-1/2})}$
where $C$ is a constant independent of the specific conductivities.
Construction of $d$ -Dimensional Zernike Bases: The authors explicitly construct and analyze Zernike bases for arbitrary spatial dimensions $d \geq 2$ , extending previous work limited to $d=2$ or $d=3$ .
Explicit Stability Constants: For the Zernike basis, the stability constant is explicitly derived as proportional to $\sqrt{|\partial B| / (4k+d)}$ , showing that stability improves as the radial index $k$ increases, provided the coefficients decay sufficiently fast.

4. Main Results

Corollary 1.2: Establishes that if $\gamma_1 - \gamma_2$ satisfies a weak maximum principle (e.g., is harmonic), Lipschitz stability holds. This serves as a foundational observation.
Theorem 1.6 & Corollary 1.7: Proves Lipschitz stability for conductivities expanded in the general basis $\{\psi_{n,k}\}$ under the coefficient decay and non-cancellation condition (1.2).
Theorem 2.2: Provides fundamental properties of the constructed radial Zernike polynomials, including their orthonormality, explicit normalization constants, and boundary values.
Application to Zernike Bases: The paper demonstrates that the $d$ -dimensional Zernike basis satisfies all necessary assumptions. Consequently, any conductivity difference expanded in this basis with coefficients satisfying the specific decay/cancellation constraints exhibits Lipschitz stability.

5. Significance and Implications

Theoretical Breakthrough: This work challenges the prevailing view that infinite-dimensional inverse conductivity problems are inherently only logarithmically stable. It demonstrates that by restricting the structure of the unknown (specifically limiting radial freedom while allowing full angular freedom), one can recover strong (Lipschitz) stability.
Reconstruction Algorithms: Lipschitz stability is crucial for the convergence guarantees of iterative reconstruction algorithms (e.g., Newton-type methods). This result suggests that reconstruction methods using Zernike expansions could converge linearly for a much broader class of conductivities than previously thought possible.
Generalization: The construction of $d$ -dimensional Zernike bases extends the utility of these functions beyond optics and tomography (where they are common in 2D/3D) to general inverse problems in higher dimensions.
Practical Relevance: The condition (1.2) implies that while the angular structure can be complex (infinite-dimensional), the radial profile must be "smooth" or "monotonic" in a specific sense (limited cancellation). This aligns with physical intuition where conductivities often vary smoothly or have specific structural constraints.

In summary, the paper bridges the gap between finite-dimensional stability and general ill-posedness by identifying a rich, infinite-dimensional class of conductivities (based on Zernike expansions) where the inverse problem is well-posed in the Lipschitz sense.