Breast tumors do not grow in isolation. They develop within a complex structural environment known as the tumor microenvironment, where cancer cells interact with blood vessels, immune cells and the extracellular matrix that surrounds them. Among the most important components of that matrix is collagen, a fibrous protein that provides tissue with mechanical strength but can also influence how cancer spreads. A new study published in Scientific Reports describes a model-free, multiparametric approach for analyzing second-harmonic generation, or SHG, microscopy images, revealing collagen patterns that may serve as distinctive signatures of breast cancer. The work by R. Mercatelli, G.M. Bernava, T. Triulzi and colleagues brings attention to a biological feature that is often visible in tissue images but difficult to measure consistently: the organization of collagen fibers around malignant tissue.
SHG microscopy is especially well suited to studying collagen because certain non-centrosymmetric structures can generate light at exactly twice the frequency of the incoming laser beam. Unlike conventional fluorescence imaging, SHG does not require the collagen to be labeled with a fluorescent dye. A pulsed laser illuminates the tissue, and highly ordered collagen fibers produce a characteristic optical signal that can be collected to form a detailed image. The resulting maps can reveal not only how much collagen is present, but also how fibers are oriented, how densely they are packed and how their local organization changes from one region to another. These characteristics matter because collagen is not merely a passive scaffold. Its architecture can affect tissue stiffness, cell migration and the physical routes available to invading tumor cells.
The challenge is that collagen organization in cancer is highly heterogeneous. Fibers may appear straight and aligned in one area, loose and disordered in another, or concentrated along the boundary between a tumor and surrounding tissue. Two samples can contain similar amounts of collagen while displaying very different spatial arrangements. Conventional image analysis often focuses on a limited number of predefined measurements or relies on classification models trained using assumptions about what a “typical” cancer pattern should look like. Such strategies can be useful, but they may overlook unexpected structures and can become difficult to generalize across patients, laboratories and imaging systems. The study’s model-free strategy addresses this problem by analyzing the information contained in SHG images without requiring the investigators to impose a rigid biological template in advance.
In practical terms, a multiparametric analysis treats each image as a collection of measurable features rather than as a single visual impression. Parameters may describe signal intensity, fiber orientation, angular dispersion, local texture, spatial distribution and the degree of alignment between neighboring structures. When considered together, these measurements create a multidimensional description of the collagen network. A model-free framework can then search for recurring patterns within that data, allowing samples to be grouped or compared according to their intrinsic image characteristics. This is different from asking an algorithm to recognize only a previously defined pattern. Instead, the data help expose which combinations of collagen properties repeatedly appear in association with breast cancer tissue.
The significance of this approach lies in the biological information encoded in collagen architecture. During tumor development, the extracellular matrix can be remodeled by cancer cells and by stromal cells such as fibroblasts. New collagen may be deposited, existing fibers may be stretched or reorganized, and bundles may become aligned around the tumor. These changes can modify the mechanical properties of the tissue and influence signaling between cells. An aligned collagen network, for example, may provide directional tracks along which tumor cells can move. Increased matrix stiffness can also alter cellular behavior through mechanosensitive pathways, potentially affecting proliferation, survival and invasion. By converting these structural changes into quantitative signatures, SHG image analysis may help connect microscopic tissue architecture with clinically relevant cancer biology.
A major advantage of the reported framework is that it could reduce the gap between sophisticated microscopy and reproducible measurement. SHG images are rich in detail, but their interpretation can depend heavily on the observer. A pathologist or researcher may recognize unusual collagen arrangements visually, yet translating that judgment into a numerical and comparable result is more difficult. Multiparametric analysis offers a way to standardize the description of the extracellular matrix. It may also help distinguish patterns that look similar to the human eye but differ in subtle features such as fiber directionality or local heterogeneity. Because the method is model-free, it has the potential to identify signatures that were not anticipated when the analysis began, an important feature when studying a disease as diverse as breast cancer.
The findings may eventually support more refined approaches to tumor classification. Breast cancer is not a single disease but a collection of molecular and pathological subtypes with different growth patterns, treatment responses and risks of recurrence. The surrounding matrix may add another layer of information to these established classifications. Collagen signatures could potentially complement tumor-cell markers, genomic data and conventional histopathology, helping researchers understand why tumors with similar cellular features can behave differently. They might also be useful for studying the tumor margin, where interactions between malignant cells and the surrounding tissue are particularly intense. However, an imaging signature should not automatically be interpreted as a clinical biomarker. It must be tested in larger, independent patient groups and evaluated against outcomes such as treatment response, recurrence and survival.
The technology also has important technical limitations. SHG signals depend on the optical properties of the tissue, the characteristics of the laser, the microscope configuration and the way images are processed. Tissue preparation can alter fiber appearance, while differences in section thickness or imaging depth may affect measured intensity and texture. A model-free analysis can reduce assumptions about biological patterns, but it does not eliminate the need for careful standardization. Researchers must still determine which parameters are robust, how reproducible they are between instruments and whether the patterns remain detectable in routine clinical specimens. In addition, collagen is only one component of the tumor microenvironment. A complete understanding of breast cancer progression will require integrating SHG measurements with information about cells, blood vessels, immune activity and molecular alterations.
Even with these cautions, the study highlights a rapidly expanding role for quantitative microscopy in cancer research. Instead of treating an image as a photograph to be inspected, scientists are increasingly using it as a source of high-dimensional biological data. The model-free multiparametric analysis of SHG images described by Mercatelli, Bernava, Triulzi and colleagues demonstrates how the physical organization of collagen can be examined systematically and converted into recognizable signatures. If future studies confirm that these signatures are linked to specific forms of tumor behavior, the approach could help researchers map the structural language of the breast cancer microenvironment. For now, its most immediate contribution is methodological: it offers a way to study collagen architecture with fewer preconceived assumptions, potentially uncovering patterns that conventional analysis leaves hidden.
Subject of Research: Collagen architecture and extracellular-matrix signatures in breast cancer analyzed through second-harmonic generation imaging.
Article Title: Model-free multiparametric analysis of SHG images reveals collagen signatures in breast cancer.
Article References: Mercatelli, R., Bernava, G.M., Triulzi, T. et al. “Model-free multiparametric analysis of SHG images reveals collagen signatures in breast cancer.” Scientific Reports (2026). https://doi.org/10.1038/s41598-026-62801-y
Image Credits: AI Generated
DOI: 10.1038/s41598-026-62801-y
Keywords: breast cancer, collagen, extracellular matrix, second-harmonic generation microscopy, SHG imaging, model-free analysis, multiparametric image analysis, tumor microenvironment.
Tags: Breast cancer collagen signaturescollagen fiber patterns in breast tumorscollagen organization as a cancer diagnostic markerextracellular matrix role in breast cancer progressionlabel-free imaging of collagen in cancer tissuesmodel-free analysis of collagen organizationmultiparametric SHG microscopy for tumor microenvironmentnon-invasive collagen imaging techniquesoptical signatures of malignant tissuesecond-harmonic generation imaging in cancer diagnosisstructural biomarkers of breast cancer using SHGtissue microarchitecture analysis with SHG microscopy
