🔬 physics

Thermal-signature equivalence of breast tumors with heterogeneous perfusion in a modified Pennes bioheat model

This study utilizes a modified Pennes bioheat model to demonstrate that while heterogeneous intratumoral perfusion patterns create distinct internal temperature distributions, heat diffusion and thermal screening often render these differences indistinguishable at the breast surface, particularly for deep-seated tumors, thereby highlighting a fundamental limitation in uniquely identifying internal tumor physiology from static thermal signatures.

Original authors: Roni Muslim, Ramacos Fardela, Tista Artu Indra Kusuma

Published 2026-05-08

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Roni Muslim, Ramacos Fardela, Tista Artu Indra Kusuma

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

Imagine your breast tissue is like a thick, multi-layered blanket made of different fabrics: a thin skin layer, a fluffy fat layer, a dense glandular layer, and a deep muscle layer. Inside this blanket, there is a small, warm "hot spot" (a tumor).

This paper asks a very specific question: If we look at the warmth on the outside of the blanket, can we tell exactly what the "hot spot" looks like on the inside?

The researchers found that the answer is often no. Here is the breakdown of their discovery using simple analogies:

1. The "Muddy Water" Analogy

Think of the tumor's internal blood flow (perfusion) as different patterns of stirring a pot of soup.

Scenario A: You stir the soup evenly.
Scenario B: You stir only the edges, leaving the center still.
Scenario C: You stir the center but leave the edges still.

Inside the pot, these three scenarios create very different swirling patterns. However, if you look at the steam rising from the top of the pot (the skin surface), the steam looks almost identical for all three. The thick layers of the pot and the air above it act like a filter, smoothing out the messy details of the stirring so that the steam just looks like "a hot pot."

The paper calls this "Thermal-Signature Equivalence." It means that two tumors with completely different internal structures can look exactly the same on the skin's surface.

2. The "Deep vs. Shallow" Rule

The researchers tested how deep the tumor sits under the skin.

The Analogy: Imagine shouting from the bottom of a swimming pool. If you are right at the surface, your voice is clear and distinct. If you are at the bottom, the water absorbs and muddles your voice. By the time the sound reaches the surface, it's hard to tell who was shouting or what they were saying; it just sounds like "noise."
The Finding: The deeper the tumor is, the more the surrounding tissue "muddles" the heat signal. A deep tumor's internal details are almost impossible to distinguish from the surface, even if the inside is very complex.

3. The "Big vs. Small" Rule

They also looked at the size of the tumor.

The Analogy: A small candle flame hidden under a thick blanket might just make a tiny, indistinct warm spot on the fabric. But a large bonfire under the same blanket will create a huge, obvious hot area.
The Finding: Bigger tumors create a stronger, more obvious heat signal. Because the signal is stronger, it's slightly easier to tell if the inside is different (like knowing the bonfire is bigger than the candle). Small tumors, however, are easily "hidden" by the tissue layers, making their internal details invisible to surface thermography.

4. The "Shape" Factor

They also checked if the shape of the breast (whether it's perfectly round or slightly deformed) changed the results.

The Finding: While the shape of the breast does change the heat pattern slightly (like how the shape of a drum affects the sound), it is a minor effect compared to the depth and size of the tumor. The "muffling" effect of the tissue depth is the dominant factor.

The Big Conclusion

The paper concludes that static thermal imaging (taking a picture of the heat on the skin) has a fundamental limit.

Just because you see a warm spot on the skin, you cannot assume you know the "personality" of the tumor underneath.

A tumor with a dead center and a hot rim might look the same as a tumor with a hot center and a cool rim.
The tissue acts like a low-pass filter (a noise-canceling headphone for heat), smoothing out the complex internal details before the heat ever reaches the skin.

In short: You can tell that there is a problem (a thermal anomaly) by looking at the skin, but you cannot reliably tell exactly what kind of problem it is (the specific internal blood flow pattern) just by looking at that heat map. The internal details get lost in the "noise" of the tissue layers.

Technical Summary: Thermal-signature equivalence of breast tumors with heterogeneous perfusion in a modified Pennes bioheat model

Problem Statement
Breast thermography offers a non-invasive method for detecting tumors by observing surface temperature anomalies caused by metabolic activity and abnormal blood perfusion. However, a critical limitation in inverse thermal modeling remains: while surface thermograms can indicate the presence of a lesion, it is unclear to what extent these surface patterns uniquely identify the internal physiological structure of the tumor. Specifically, real tumors often exhibit spatially heterogeneous perfusion (e.g., viable rims, necrotic cores, anisotropic growth), yet standard models often assume uniform perfusion. The central question addressed is whether different intratumoral perfusion structures, which create distinct internal temperature fields, produce distinguishable thermal signatures at the breast surface, or if heat diffusion and tissue properties render them effectively indistinguishable.

Methodology
The authors employ a modified Pennes bioheat equation to model steady-state heat transfer in multilayer breast tissue (skin, fat, glandular tissue, and muscle) containing a finite-sized tumor.

Model Formulation: The tumor is treated as a finite circular inclusion with spatially heterogeneous perfusion rates ( $\omega_t(\mathbf{x})$ ), contrasting with the classical assumption of uniform perfusion. Several perfusion profiles are simulated: uniform, rim-enhanced, necrotic-core, and anisotropic.
Geometry: The model utilizes a two-dimensional domain representing a cross-section of the breast, bounded by a chest wall (Dirichlet condition) and an outer surface (Robin condition for convective heat loss). The geometry allows for both idealized flat surfaces and weakly deformed surfaces.
Quantitative Framework: Instead of focusing solely on internal temperature fields, the study introduces a "thermal-signature equivalence" framework. It defines a profile-based $L_2$ distance ( $d_{L2}$ ) between surface temperature rise profiles ( $\Delta T_s(y)$ ) generated by different perfusion configurations.
Equivalence Criterion: Two tumor configurations are deemed "thermally equivalent" at the surface if their profile distance falls below a specific observational tolerance ( $\varepsilon_T$ ). This metric quantifies the limit of distinguishability.
Numerical Implementation: The governing equations are solved using finite-difference/finite-volume discretization on a structured grid. Mesh convergence tests ensure results are independent of numerical resolution. The study compares two parameter sets (Khomsi-based and Lozano-inspired) to test robustness.

Key Results

Internal vs. Surface Distinguishability: While different perfusion patterns (uniform, rim-enhanced, necrotic-core, anisotropic) generate clearly distinct internal temperature distributions, these differences are significantly smoothed by heat diffusion and thermal screening as they propagate to the surface. Surface profiles for different internal structures often appear nearly identical.
Thermal-Signature Equivalence: The study demonstrates that tumors with fundamentally different internal vascular organizations can produce surface signatures that are indistinguishable within realistic observational tolerances. The $d_{L2}$ metric shows that many perfusion pairs collapse into the same "thermal class" at the surface.
Impact of Tumor Depth: Tumor depth is identified as a primary factor reducing distinguishability. As depth increases, the profile distance between different perfusion patterns decreases exponentially. Deeper tumors undergo stronger thermal attenuation, making their internal heterogeneity increasingly difficult to resolve from static surface data.
Impact of Tumor Size: Conversely, increasing tumor diameter enhances distinguishability. Larger tumors generate a greater total thermal perturbation, making differences in perfusion patterns more visible at the surface. The relationship between profile distance and tumor diameter follows an approximate power law ( $d_{L2} \propto D^b$ ).
Secondary Factors: Fat-layer thickness and outer geometry deformation modulate the surface signature but act as secondary factors compared to depth and size. Fat thickness does not act as a simple monotonic attenuator but influences heat redistribution in complex ways.

Significance and Claims
The paper claims to highlight a fundamental limitation of static breast thermography: the detection of a thermal anomaly does not guarantee the unique identification of intratumoral perfusion heterogeneity.

Non-Uniqueness: The study establishes that surface thermograms are not direct maps of internal vascular structures but are filtered projections. Consequently, the similarity of surface thermograms should not be interpreted as similarity in internal physiology, nor should the inability to distinguish two surface signatures imply identical internal structures.
Operational Framework: The introduced "thermal-signature equivalence" framework provides a quantitative tool to assess when different internal states become indistinguishable, bridging the gap between lesion detectability and structural identifiability.
Modeling Implications: The results suggest that inverse thermal modeling efforts aiming to reconstruct detailed internal perfusion maps from static surface data face intrinsic physical limits imposed by diffusive smoothing.

The authors conclude that while thermography is valuable for detecting anomalies, its ability to characterize internal tumor heterogeneity is constrained by the physics of heat transfer. Future work suggested by the authors includes extending the framework to three-dimensional models, patient-specific geometries, dynamic thermography, and clinical validation.

1. The "Muddy Water" Analogy

2. The "Deep vs. Shallow" Rule

3. The "Big vs. Small" Rule

4. The "Shape" Factor

The Big Conclusion

More like this