A Nearable Soft Mat Based on Distributed Optical Fiber Sensing for Physiological Monitoring

arXiv:2607.11255 · cs.CV, physics.app-ph, physics.optics · Submitted 2026-07-13 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: AI Radio. Generated commentary on the latest Artificial Intelligence papers.

Tom: We're looking at the paper: "A Nearable Soft Mat Based on Distributed Optical Fiber Sensing for Physiological Monitoring".

Jane: Here's the gist of the paper. Distributed optical fiber sensing (DOFS) enables high-resolution distributed measurements using optical frequency domain reflectometry (OFDR) to create a soft mat interface for nearable physiological monitoring This study presents a soft DOFS-based mat as a large-area interface for physiological monitoring The system successfully demonstrated the feasibility of combining large-area distributed sensing, spatial mapping, and quantitative cardiorespiratory monitoring within a DOFS-based soft nearable interface.

Tom: System Design and Fabrication The proposed system consists of a single-mode optical fiber embedded in a soft silicone matrix and arranged in a serpentine layout to cover an extended sensing area.

Jane: The mat has a square geometry of 40 cm × 40 cm, a total thickness of 1 cm, and the optical fiber is embedded at mid-thickness.

Tom: DragonSkin TM 20 silicone rubber was selected because of its biocompatibility, flexibility, conformability to non-planar surfaces, and ability to provide mechanical protection to the embedded fiber.

Paper summary: Tom: Okay, so we’re talking about a paper called "A Nearable Soft Mat Based on Distributed Optical Fiber Sensing for Physiological Monitoring". Basically, the authors are showing how distributed optical fiber sensing, or DOFS, can be used with optical frequency domain reflectometry, OFDR, to create a soft interface for monitoring things like our heart and breathing.

Jane: So what's the core idea here? What are they actually claiming they can do?

Tom: They’re proposing this soft mat as a large area interface for physiological monitoring. It claims they can combine large area distributed sensing with spatial mapping and quantitative cardiorespiratory monitoring all within this DOFS-based soft nearable interface.

Lu: I'm interested in the idea of combining those things, because usually, when you have distributed sensing, you either get a lot of data points that are hard to map spatially or you have a very localized measurement that doesn't give you much context about what’s happening across the whole area.

Jane: Right. So they’re tackling the need for sensitive and spatially resolved interfaces in biomedical applications, which is where fiber optic sensors usually shine, but they say it hasn't been explored enough for this kind of monitoring yet.

Meng: I wonder about the physical setup here because fiber optics are fragile, right? They have to protect that optical fiber while still keeping it flexible enough to conform to surfaces and still letting mechanical changes get measured effectively.

Tom: Exactly, and they used a specific material called DragonSkin TM twenty silicone rubber for the mat <ref:2607.11255#pg1>. The fiber is embedded in this soft silicone matrix in a serpentine layout across a forty by forty centimeter area, with the fiber placed right at the mid-thickness of that one centimeter thick mat <ref:2607.11255#pg1>.

Jane: That serpentine layout is key because it maximizes how much sensing coverage they get over that surface area.

Lu: And they chose DragonSkin TM twenty silicone rubber because it has biocompatibility and it can conform to non-planar surfaces, which is pretty important for a soft interface, and it also protects the fiber from mechanical damage <ref:2607.11255#pg1>.

Meng: But what about the sensing mechanism itself? How do they actually get the physiological data out of that physical setup?

Jane: That comes down to Rayleigh backscattering, which means they are looking at microscopic changes in the refractive index along the fiber core. When something external, like a change in strain or temperature, messes with that fiber, it shifts the Rayleigh signature compared to a reference state.

Tom: So they discretize this fiber into sensing sites by setting a gage pitch of two point six millimeters over an active area of forty by forty centimeters, which gives them two thousand two hundred fifty sensing sites at a sampling frequency of fifty hertz.

Lu: That discretization defines the spatial sampling step, and they’re using that to define their spatial resolution. If you look at the math, that means the error in your measurement falls with the square of the sampling interval one <ref:2607.11255#pg1>.

Paper summary: Meng: So, what are they doing with all those two thousand two hundred fifty sites? Are they just dumping a huge stream of raw data there?

Jane: Not exactly. They process these distributed signals to reconstruct time-resolved two dimensional response maps that show how the subject interacts with the surface mechanically. From there, they extract respiratory and cardiac components to estimate the respiratory rate and heart rate.

Tom: How do they actually calculate those rates from all that data?

Jane: The respiratory rate is calculated using a formula: sixty divided by the difference between two consecutive time points multiplied by one apm. And for the heart rate, they use the Welch method in the frequency domain to compute the power spectral density.

Lu: That sounds like a standard approach for signal processing, but how does that process handle the noise inherent in sensing through a soft tissue interface?

Tom: They show that when they process it this way, all subjects and conditions showed a mean absolute error below one apm for respiratory rate estimation, except for subject five.

Meng: That’s a tight error bound if you're dealing with moving subjects. What about the heart rate estimation accuracy across different conditions?

Jane: The heart rate estimation also showed good accuracy across all subjects and conditions, with mean absolute error values staying below three point two one bpm.

Lu: So they found that this distributed measurement approach offers a way to identify specific sensing sites or spatial regions that provide the highest quality respiratory and cardiac signals.

Tom: That’s the key implication for me. It suggests that instead of just looking at the average signal, you can pinpoint exactly where on that mat is giving you the most reliable physiological information.

Meng: So what does this mean for deployment? If we use this to monitor someone sitting down, how much does it actually change compared to a standard belt or chest strap?

Jane: The pilot assessment showed that even in a challenging seated configuration, they could estimate both the respiratory rate and the heart rate in agreement with their reference system.

Lu: But you have to remember what the paper flags as a limitation—their current signal-processing approach is an initial strategy, and they haven't fully exploited all twenty two hundred fifty sensing sites yet.

Tom: So it’s not perfect yet, but it proves that this distributed sensing method can work for monitoring physiology on a soft interface.

Meng: For us in production, the question is how much computational cost does running this large distributed system add to the overall monitoring solution?

Jane: That’s what we need to figure out next. But for now, the main point of this paper is that combining these sensing elements gives us a more comprehensive view of cardiorespiratory signals than a single measurement could provide.

Conclusion: Tom: So we’ve been looking at how they took distributed optical fiber sensing, or DOFS, and put it on a soft mat to monitor heart rate and breathing in real life.

Jane: Right, so this paper is about using that fiber setup—the distributed sensing—to create a soft interface for physiological monitoring.

Tom: They built this whole system up using a single mode optical fiber embedded in DragonSkin TM twenty silicone rubber, which they laid out in a serpentine pattern over a forty by forty centimeter area.

Lu: The main idea is combining that large-area distributed sensing with spatial mapping to get quantitative cardiorespiratory monitoring all at once on this soft mat.

Meng: It sounds like they’re trying to solve the problem of getting high-quality, spatially resolved physiological data from a flexible surface, which is tough because things move around.

Lalam: The researchers successfully demonstrated that you can combine large-area distributed sensing with spatial mapping to monitor cardiorespiratory signals on this soft interface.

Tom: What does that actually mean for us in terms of what this technology can do?

Jane: It means they proved the feasibility of combining large area distributed sensing, spatial mapping, and quantitative cardiorespiratory monitoring within a DOFS-based soft nearable interface.

Lu: They showed that their system successfully demonstrated the feasibility of combining large area distributed sensing with spatial mapping and quantitative cardiorespiratory monitoring within this soft nearable interface.

Meng: So they proved it works in principle, but what’s the real-world cost or complexity when you try to actually ship this thing out?

Tom: The results show that all subjects and conditions showed a mean absolute error below one apm for respiratory rate estimation, except for subject five.

Jane: And the heart rate estimation also showed good accuracy across all subjects and conditions, with mean absolute error values remaining below three point two one bpm.

Lu: The distributed nature of this measurement offers the possibility of overcoming limitations by identifying the sites or spatial regions providing the highest quality respiratory and cardiac signals.

Tom: The pilot assessment also showed that this distributed measurement could estimate both respiratory rate and heart rate in agreement with their reference system, even when the subject was in a challenging seated configuration.

Jane: But they also flagged a limitation—this signal-processing approach is just an initial strategy; they haven't fully exploited all two thousand two hundred fifty sensing sites yet.

Lu: The authors suggest that future work needs to focus on spatially adaptive processing strategies capable of identifying those specific sensing sites with the highest physiological information content.

Tom: So, while it works, they aren't quite at their peak performance level yet regarding how much spatial detail they can actually extract from all those measurement points.

Jane: It’s a big step because it shows that this distributed sensing approach is a viable way to get comprehensive data for physiological monitoring on soft surfaces.

Lu: This paper sets the stage for moving towards spatially adaptive processing, which is where the real power of this type of large-scale sensing comes from.

V. Lavorgna, M. Pulcinelli, A. Polimadei, R. D’Amato, C. Massaroni, M.A. Caponero, E. Schena, D. Lo Presti

Unit of Measurements and Biomedical Instrumentation, Department of Engineering, Università Campus Bio -Medico di Roma · Photonics Micro- and Nanostructures Laboratory, ENEA Research Center of Frascati · Fondazione Policlinico Universitario Campus Bio -Medico

cs.CV, physics.app-ph, physics.optics

Submitted: 2026-07-13

Updated: 2026-07-13

Journal ref: V. Lavorgna et al., "A Nearable Soft Mat Based on Distributed Optical Fiber Sensing for Physiological Monitoring," in IEEE Sensors Journal, vol. 26, no. 20, pp. 30429-30437, 15 Oct.15, 2026

DOI: 10.1109/JSEN.2026.3727664

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 67/100

The gist: Distributed optical fiber sensing (DOFS) enables high-resolution distributed measurements using optical frequency domain reflectometry (OFDR) to create a soft mat interface for nearable physiological…

Key concepts

Distributed Optical Fiber Sensing (DOFS)
This technique uses an optical fiber embedded in a soft material to measure changes across a large surface. It works by detecting microscopic changes in the fiber's light properties, which are caused by external factors like breathing or heartbeats. This allows for high-resolution monitoring over an extended area.
Optical Frequency Domain Reflectometry (OFDR)
OFDR is the specific method used to measure the signals from the optical fiber. It analyzes how light reflects off microscopic changes along the fiber's core. By measuring these reflections, researchers can determine local changes in refractive index caused by physiological movements.
Rayleigh Backscattering
This is the physical phenomenon that forms the basis of sensing. Rayleigh backscattering occurs when light interacts with tiny refractive index fluctuations along the fiber core. When a subject moves or breathes, these fluctuations change, causing a measurable shift in the reflected optical signal.

Terminology

Summary

Distributed optical fiber sensing (DOFS) enables high-resolution distributed measurements using optical frequency domain reflectometry (OFDR) to create a soft mat interface for nearable physiological monitoring This study presents a soft DOFS-based mat as a large-area interface for physiological monitoring The system successfully demonstrated the feasibility of combining large-area distributed sensing, spatial mapping, and quantitative cardiorespiratory monitoring within a DOFS-based soft nearable interface. p. 0

System Design and Fabrication

The proposed system consists of a single-mode optical fiber embedded in a soft silicone matrix and arranged in a serpentine layout to cover an extended sensing area. The mat has a square geometry of 40 cm × 40 cm, a total thickness of 1 cm, and the optical fiber is embedded at mid-thickness. p. 3 DragonSkin TM 20 silicone rubber was selected because of its biocompatibility, flexibility, conformability to non-planar surfaces, and ability to provide mechanical protection to the embedded fiber. p. 3 The optical fiber was arranged in a continuous serpentine layout to maximize the sensing coverage over the mat surface. p. 3

Sensing Mechanism and Data Acquisition

The sensing mechanism relies on Rayleigh backscattering, which originates from microscopic refractive index fluctuations along the fiber core. When the fiber is subjected to external perturbations, such as mechanical strain or temperature variations, the local optical properties of the fiber change, producing a shift of the Rayleigh signature with respect to a reference state. p. 3 The discretization of the fiber into distributed sensing sites defines the spatial sampling step, or gage pitch. p. 3 With a gage pitch of 2. p. 06 mm, the system provided 2250 sensing sites across the active area at a sampling frequency of 50 Hz. p. 0

Data Processing and Physiological Estimation

The analysis first investigates the mechanical interaction between the subject and the sensing surface through the reconstruction of time-resolved two-dimensional response maps. The distributed signals are processed to extract respiratory and cardiac components and to estimate respiratory rate (RR) and heart rate (HR). p. 2 RR is computed using the formula RR = 60 / (t i+1 - t i) [apm]. p. 7 HR is estimated in the frequency domain using the Welch method to compute the power spectral density (PSD). p. 8

Performance and Conclusion

The results show that all subjects and conditions showed an MAE below 1 apm for respiratory rate estimation, with the exception of subject S5. The heart rate estimation also showed good accuracy across all subjects and conditions, with MAE values remaining consistently below 3. p. 021 bpm. The distributed nature of the DM offers the possibility of overcoming limitations by identifying the sites or spatial regions providing the highest-quality respiratory and cardiac signals. p. 2 Future work will focus on spatially adaptive processing strategies capable of identifying the sensing sites with the highest physiological information content. p. 1 The pilot assessment also showed that the DM could estimate both RR and HR in agreement with the BH reference system despite the challenging seated configuration. p. 2 The present signal-processing approach should nevertheless be considered an initial strategy that does not yet exploit the full spatial information provided by the 2250 sensing sites. p. 0

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI system can perform spatially resolved physiological monitoring by reconstructing time-resolved two-dimensional maps of mat output, which reflecting back–mat mechanical coupling and cardiorespiratory-induced perturbations. This allows for an analysis of individual respiratory mechanics and identifying possible asymmetries or abnormal ventilatory patterns.

  2. Improved AI system can estimate respiratory rate (RR) and heart rate (HR) by utilizing a strategy based on the spatial averaging of selected sensing sites, which is noted as a way to extract robust global signals representative of the dominant respiratory and cardiac dynamics. This approach allows for reliable estimation even in suboptimal postural conditions or when the back–mat contact is limited to a small portion of the sensing surface.

  3. Improved AI system can generate two-dimensional response maps that correlate distributed DM output with physiological reference signals from the BioHarness, allowing for the qualitative interpretation of how respiratory and cardiac activity modulates the mechanical interaction between the subject’s back and the sensing surface. This capability supports identifying regions where heartbeat-induced mechanical components are more evident, even though it requires further signal processing to isolate cardiac components.

Related papers