🧬 biology

Chirality is invisible to displacement data: a general symmetry-group criterion and taxonomy for the identifiability of active tissue deformations

This paper establishes a general, constitutive symmetry-group criterion that determines whether active tissue deformations, particularly chiral (rotational) components, are identifiable from displacement data by proving that such deformations are indistinguishable up to a rotation if their parametrized class intersects with its own orbit under the energy-invariance group $SO(d)$ .

Original authors: Rachele Allena

Published 2026-09-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Rachele Allena

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Living tissues are not passive bags of cells waiting to be stretched by external forces; they are active, living machines that constantly reshape themselves from the inside. Cells divide, elongate, and twist, pushing and pulling on their neighbors to drive the complex folding of a brain, the looping of a gut, or the twisting of a developing heart. Scientists have long used mathematical models to describe this process, separating the shape change caused by the tissue's own internal activity from the elastic stretching that happens in response. A key part of these models is a mathematical object representing the active, internal drive of the tissue. The challenge for researchers has been to work backward: if they can measure how a piece of tissue has moved and deformed, can they figure out exactly what internal twist or rotation caused that movement?

This question sits at the intersection of biology and physics, specifically in a field that treats growing tissue like a rubber sheet that is being actively reshaped. While it is relatively straightforward to predict how a tissue will move if you know its internal activity, the reverse problem is notoriously difficult. In many cases, different internal forces can produce the exact same external shape, making it impossible to know which one actually happened. A recent study by Rachele Allen at the University of Côte d'Azur tackles this ambiguity head-on, moving beyond trial-and-error computer searches to find a fundamental rule that explains when and why certain internal twists remain invisible to observation.

The researchers focused on a specific type of internal drive called chirality, which is the property of having a "handedness," like a left hand versus a right hand. In biological development, this handedness is crucial; for instance, the human heart must twist in a specific direction to function correctly. The central mystery was whether scientists could ever tell, just by looking at the final position of the tissue, whether the internal drive was a left-handed twist or a right-handed one. Previous computer experiments had hinted that for certain types of twisting, the answer was no, but those results were based on checking specific examples one by one, leaving open the possibility that other types of twists might behave differently.

Allen's work provides a definitive, general answer by looking at the underlying symmetry of the tissue's energy. The study proves that for a wide range of common biological materials, the tissue's internal energy does not care about the direction of a rigid rotation. If you take a tissue and rotate its internal driving force by a certain angle, the resulting physical movement of the tissue remains exactly the same. This means that a left-handed twist and a right-handed twist can produce identical displacement patterns, making them indistinguishable to an observer who only sees the final shape. The paper establishes a clear, geometric test to determine if a specific type of internal drive can be identified: if rotating the internal drive keeps it within the same category of possible drives, then it cannot be uniquely identified.

The researchers applied this new rule to several different mathematical categories of tissue behavior. They confirmed that for tissues driven by simple stretching or symmetric shearing, the internal forces can be identified with high precision. However, for tissues driven by a pure twist, the rule predicts a fundamental ambiguity. The study showed that a left-handed twist and a right-handed twist are mathematically indistinguishable in this framework; the data simply cannot tell them apart. This explains why previous computer searches struggled to find the correct "handedness" for these specific cases. The researchers did not just rely on theory; they built a computer simulation of a ring-shaped piece of tissue and solved the physics equations twice: once with a left-handed twist and once with a right-handed twist. The resulting movements were identical down to the limits of computer precision, proving that the ambiguity is a real physical feature, not just a mathematical quirk.

Crucially, the study also tested this rule on a new, more complex type of tissue behavior that had never been analyzed before. Without knowing the answer in advance, the researchers used their geometric rule to predict exactly where the ambiguity would occur. They found that in this larger, more complex category, the inability to identify the internal force was confined to a single, specific line of twisting behavior, while all other variations remained identifiable. This prediction was later confirmed by a massive computer scan of millions of possibilities, which found that the rule correctly identified the exact spot where the data failed to distinguish between different internal drives. This demonstrates that the rule is a reliable tool that can predict identifiability without needing to run expensive and time-consuming computer searches for every new scenario.

The findings suggest that to fully understand how tissues twist and turn, scientists cannot rely on shape data alone. Because the tissue's energy is indifferent to certain rotations, the "handedness" of the internal drive is effectively hidden from view if only the final shape is measured. To resolve this, future studies would need to combine shape data with other types of information, such as tracking the orientation of specific cell clusters over time, to break the symmetry and reveal the true direction of the twist. By establishing a clear boundary between what can and cannot be known from displacement data, this work provides a new foundation for understanding the mechanical secrets of life's most complex shapes.

Technical Summary: Chirality is Invisible to Displacement Data

Problem Statement
Biological tissues undergo active deformations driven by cellular processes (proliferation, elongation, reorientation) that often include a rotational or chiral component. In the context of morphoelasticity, the deformation gradient $F$ is multiplicatively decomposed into an active tensor $F_a$ and an elastic tensor $F_e$ . A critical inverse problem arises: can the active tensor $F_a$ be uniquely identified from observations of the resulting displacement field $u$ ?

Previous work (specifically a companion paper by the same author) demonstrated a dichotomy in two-dimensional (2D) identification problems. While isotropic, symmetric anisotropic, and symmetric shear classes of $F_a$ were identifiable to near machine precision, a skew-symmetric class exhibited a fundamental, irreducible sign ambiguity (i.e., $F_a(s)$ and $F_a(-s)$ produced identical displacements). However, these findings were established through case-by-case numerical searches and specific algebraic identities, leaving open whether the ambiguity was an artifact of specific parametrizations or a general structural property of the forward problem. Furthermore, there was no general criterion to predict identifiability for untested classes of active tensors.

Methodology
The paper reformulates the identifiability problem not as a search over specific parameterizations, but as a question of the symmetry group of the constitutive energy density $W$ .

Forward Operator and Equivalence: The author defines an equivalence relation where two active tensors $F_a$ and $F'_a$ are observationally equivalent if they yield the same displacement field. The set of all such equivalent tensors forms the "fiber" of the forward operator.
Symmetry Group Analysis: The study identifies that the fiber of the forward operator contains the left orbit of $F_a$ under the material symmetry group of the stored-energy density. Specifically, if $Q$ is an element of the material symmetry group such that $W(F_e Q) = W(F_e)$ , then $F_a$ and $Q F_a$ are indistinguishable.
Constitutive Derivation: For the compressible Neo-Hookean energy (the constitutive law used in the companion study), the author analytically proves that the material symmetry group is exactly $SO(d)$ , the group of proper rotations. This implies that any rigid pre-rotation of the active tensor is invisible to displacement data.
Orbit-Intersection Criterion: The author derives a general criterion for class-restricted identifiability. An active tensor $F_a$ within a specific parametrized class $\mathcal{C}$ is identifiable if and only if the $SO(d)$ -orbit of $F_a$ does not re-enter the class $\mathcal{C}$ at any point other than $F_a$ itself.
Validation: The theoretical results are verified through:
- Symbolic algebraic checks (confirming the symmetry group and rotation identities).
- Monte Carlo numerical scans over random tensor classes.
- Direct nonlinear finite-element simulations solving the forward problem for $F_a$ and $Q F_a$ under non-symmetric boundary conditions to confirm identical displacement fields to machine precision.

Key Contributions

General Symmetry Criterion: The paper establishes a single, constitutive, geometry-independent criterion for identifiability. Identifiability is determined purely by whether the $SO(d)$ -orbit of a class member intersects the class elsewhere.
Unified Explanation of Ambiguity: The criterion explains the previously observed sign ambiguity in the skew-symmetric class as a consequence of the class being closed under $SO(2)$ rotations (up to reparametrization). Conversely, it proves that symmetric classes are identifiable because their orbits do not re-enter the class (except at the identity), provided the tensors are physically admissible (positive-definite).
Predictive Power: The criterion is applied prospectively to a two-parameter off-diagonal class never before tested. It correctly predicts that non-identifiability is confined strictly to the skew-symmetric line ( $p = -q$ ) within this larger class, a result confirmed by numerical scans. This demonstrates the tool's ability to identify non-identifiable loci without prior knowledge of the answer.
Taxonomy of Identifiability: The results organize disparate identification outcomes into a unified taxonomy based on constitutive symmetry rather than specific parametrization.

Results

Theoretical: For compressible Neo-Hookean materials, the material symmetry group is exactly $SO(d)$ . Consequently, $F_a \sim Q F_a$ for all $Q \in SO(d)$ .
2D Specifics:
- Identifiable Classes: Isotropic, general symmetric anisotropic, and symmetric shear classes are identifiable because no non-trivial rotation maps a member of these classes back into the class while maintaining positive-definiteness.
- Non-Identifiable Class: The skew-symmetric class is not identifiable; $F_a(s)$ and $F_a(-s)$ are observationally equivalent because a specific rotation $Q(s) \in SO(2)$ maps one to the other within the class.
- Two-Parameter Class: In the general off-diagonal class $F_a(p,q)$ , non-identifiability occurs only when $p+q=0$ (the skew-symmetric line). All other points in the class are identifiable.
Numerical: Finite-element simulations confirm that solving the forward problem for $F_a$ and $Q F_a$ yields displacement fields that coincide to machine precision ( $10^{-14}$ to $10^{-15}$ relative error), even when boundary conditions break rotational symmetry.

Significance and Claims
The paper claims to close the gap between case-specific numerical findings and general structural understanding. By shifting the focus from "searching for ambiguities" to "checking for symmetry orbit intersections," the author provides a reliable, checkable tool for determining identifiability a priori.

The significance lies in:

Moving beyond case-by-case analysis: The criterion allows researchers to determine identifiability for any new parametrized class without exhaustive numerical searching.
Clarifying the nature of chirality: It provides a precise mechanical statement that chiral active deformations are identifiable only up to their magnitude, never their handedness, when relying solely on displacement data.
Complementing existing literature: The author positions this work as a necessary analytical step before numerical reconstruction. They argue that one should first rule out intrinsic constitutive non-identifiability (via symmetry) before attempting numerical inversion.
Practical utility: The successful prediction of the non-identifiable locus in the untested two-parameter class (correcting an initial hand-derivation error) serves as concrete evidence that the criterion is a robust practical tool, not merely an abstract existence statement.

The author remains modest regarding the scope, noting that the results are specific to the compressible Neo-Hookean energy and that the criterion provides a sufficient condition for non-identifiability (via symmetry) but does not rule out other geometry-specific sources of non-uniqueness. They also note that while the 2D case yields discrete ambiguities, the 3D case ( $SO(3)$ ) would generically lead to continuous non-identifiability, a direction for future work.

Technical Summary: Chirality is Invisible to Displacement Data

More like this