A New Skin Model Could Help Make Optical Health Devices More Accurate for Every Skin Tone
Medical devices that read the body with light may soon have a more realistic way to test their performance across the full spectrum of human skin tones. Researchers at VTT Technical Research Centre of Finland have developed multilayer “optical phantoms” that mimic the appearance and light-scattering behavior of human skin while also containing artificial blood vessels through which a blood-like fluid can flow. The models are designed to give engineers a controlled, repeatable platform for evaluating technologies such as pulse oximeters, hyperspectral cameras, and wearable health sensors before those devices reach patients.
Many modern health technologies rely on a deceptively simple principle: light is directed into the skin, and sensors analyze the portion that is reflected or transmitted back. Changes in the returning signal can reveal information about blood oxygenation, pulse, blood volume, or tissue composition. Yet the path of light through skin is strongly influenced by pigmentation. Research has raised concerns that some optical devices are less accurate for people with darker skin, creating a significant challenge for healthcare equity and patient safety. Human testing remains essential, but it can be difficult to control every variable. A laboratory phantom can provide the same target again and again under precisely defined conditions.
The Finnish team constructed its models as layered silicone structures that reproduce several key regions of the body’s surface. The upper layer represents the epidermis, the outermost portion of skin, while deeper silicone layers imitate underlying tissue and subcutaneous fat. Different pigments were incorporated to create versions corresponding to lighter, medium, and darker skin tones. Because silicone can be engineered to control optical properties such as absorption and scattering, the researchers were able to tune the models so that they interacted with light in ways resembling real human skin.
The models also included a synthetic vascular network positioned beneath the skin-like layers. A miniature pump circulated a blood-mimicking liquid through the artificial vessels, allowing the researchers to reproduce an important feature missing from many static skin phantoms: movement. In living tissue, blood is constantly flowing and changing the amount of light absorbed by the skin. That dynamic behavior is critical for technologies that detect pulsation or calculate blood-related signals. By connecting the vessels to a pump, the researchers could study how pigmentation, tissue depth, vessel placement, and simulated blood flow affected the signals available to an optical sensor.
To determine whether the phantoms behaved like real skin, the team examined them with hyperspectral imaging. Unlike an ordinary camera, which records only broad red, green, and blue channels, a hyperspectral system collects data across many narrow wavelengths. The resulting spectral profile can reveal how strongly a material absorbs or reflects light at different points. The researchers compared areas overlying the artificial vessels with nearby regions without vessels, measuring whether the blood-like fluid produced a detectable signature after light had passed through the pigmented surface layers.
The results highlighted the optical problem that can arise as pigmentation increases. In the lighter models, the flowing blood-like fluid generated a clear signal that could be distinguished from surrounding tissue. In the medium and darker models, that signal became progressively harder to identify. In the darkest phantom, absorption by the upper pigmented layer largely concealed the spectral signature associated with the fluid beneath it. This does not mean that every device will fail on darker skin, but it demonstrates how pigmentation can reduce the amount of useful information reaching a sensor and make blood-related measurements more challenging.
The findings offer a physical explanation for a concern that has emerged in studies of optical medical devices. Melanin, the pigment responsible for much of the variation in human skin color, absorbs light over a broad range of wavelengths. When more light is absorbed near the surface, less reaches deeper blood vessels and less of the vessel-related signal returns to the detector. Device designers can respond by selecting different wavelengths, increasing sensor sensitivity, improving algorithms, or changing the geometry of illumination and detection. However, such solutions can only be evaluated reliably if testing materials represent the diversity of the people who will use the devices.
The new phantoms could help manufacturers and researchers identify performance gaps earlier in the development process. They provide a standardized environment in which the same sensor can be tested on several skin-tone models while blood flow, vessel depth, fluid properties, and lighting conditions are controlled. This makes it possible to separate an instrument’s technical limitations from the biological variability encountered in human studies. The models may also be useful for comparing pulse oximeters, wearable monitors, and hyperspectral systems under conditions that would be difficult or ethically complicated to reproduce repeatedly with volunteers.
Durability is another important feature for a research tool intended for long-term use. After nine months, the optical properties of the silicone phantoms had changed only slightly, suggesting that they could remain useful for extended testing in academic laboratories and industry settings. The researchers now plan to broaden the range of represented skin tones and compare measurements from the phantoms directly with data collected from human skin. Their work does not replace clinical validation, but it could make that validation more systematic and inclusive, helping optical health technologies move toward consistent performance across diverse populations.
Subject of Research: Skin-tone representation in optical medical device testing and vascular optical phantoms
Article Title: Enhancing skin tone representation in optical vascular phantoms
Web References: Journal of Biomedical Optics article; VTT Technical Research Centre of Finland
References: A. Ranta-Lassila et al., “Enhancing skin tone representation in optical vascular phantoms,” Journal of Biomedical Optics 31(7), 075001 (2026). DOI: 10.1117/1.JBO.31.7.075001
Image Credits: Alexey Popov, VTT, Sensing Solutions
Keywords
Optical phantoms, skin tone, medical devices, pulse oximeters, wearable sensors, hyperspectral imaging, biomedical engineering, optics, blood-flow simulation, healthcare equity
