Smart bandage clears new hurdle: monitors chronic wounds in human patients

Caltech engineers have taken their flexible “lab-on-skin” bandage out of the animal lab and onto the wards, logging round-the-clock data from 20 human patients with stubborn wounds that resist healing. In the first peer-reviewed report of the device’s performance in people, the researchers show that the stamp-sized patch can scoop up fresh fluid from chronic ulcers, scan it for tell-tale molecules of inflammation and wirelessly beam the results to a smartphone, often one to three days before visible symptoms appear.

The results, detailed recently in Science Translational Medicine (DOI: 10.1126/scitranslmed.adt0882), clear a key translational hurdle for a technology that could lighten the load on nurses, give doctors earlier warning of infection and, ultimately, speed recovery for millions with slow-healing injuries. The Caltech news office, which first reported the study under the headline “Smart bandage logs real-time wound data in first human study,” noted that the trial builds on successful animal work published in 2023. That earlier work demonstrated the bandage’s potential in animal models not only monitoring biomarkers indicative of infection or inflammation but also delivering medication or electrical stimulation directly to the wound site to accelerate healing.

Lead researcher Wei Gao, a professor of medical engineering at Caltech and a Heritage Medical Research Institute investigator, developed the so-called iCare smart bandage with collaborators at the Keck School of Medicine of the University of Southern California. The prototype used in the study sits on a flexible, biocompatible polymer substrate and is manufactured with low-cost, 3-D-printable techniques. Co-lead authors included Canran Wang and Kexin Fan.

Early warning system and healing prediction

The disposable layer houses a nano-engineered sensor array that zeroes in on two reactive molecules [nitric oxide (NO) and hydrogen peroxide (H₂O₂)] that spike when tissue turns inflamed or infected. A reusable printed circuit board snap-fits on top, powering the electronics and relaying data to a phone (or computer or tablet) in real time. “Microfluidics” carved into the patch guide new exudate away from the wound bed, while three nested modules—a membrane, a bio-inspired shuttle and a micropillar drain—keep the sample uncontaminated by older fluid.

“Our innovative microfluidics remove moisture from the wound, which helps with healing. They also make sure that samples analyzed by the bandage are fresh, not a mixture of old and new fluid. To get accurate measurements, we need to sample only the newest fluid at a wound site. In this way, iCares can watch in real time for important biomarkers of inflammation and infection.” —Wei Gao, Caltech in an announcement.

That freshness mattered in the 20-patient pilot, which enrolled people coping with diabetic foot ulcers, circulation-related sores and pre- or post-surgical wounds. According to the Caltech release, the microfluidic shuttle kept biomarkers stable long enough for the on-board sensors to register meaningful chemical shifts. Those shifts turned up one to three days before redness, swelling or odor appeared, giving clinicians a head start.

More than an early alarm

The patch does more than ring an early alarm. A machine-learning model bundled with the software crunches time-stamped readings and predicts how quickly the tissue will close, matching the accuracy of expert caregivers, the researchers report. Such foresight can help schedule debridement, adjust antibiotics or decide whether a patient can safely be managed at home.

Today, wound care teams still rely heavily on visual inspection and manual measurement. Those are methods that can miss fast-moving infections or subtle changes in biochemical status. The iCare approach replaces that episodic snapshot with a streaming feed, offering a fuller picture of what is happening beneath the gauze.

Optimizing moisture

Technical details aside, the device also performs a basic housekeeping task: removing excess moisture. Wounds need to stay damp enough to encourage cell migration but dry enough to avoid maceration. By siphoning off surplus exudate, the bandage helps maintain that balance and, as Gao’s group hopes, may eventually shorten healing time.

Co-author David G. Armstrong, a surgeon at USC’s Keck School of Medicine, provided clinical access and perspective on how the tool might slot into existing workflows. Although the current paper focuses on feasibility and analytical validation, the team plans larger trials that track outcomes such as time-to-closure and hospital readmission rates. The long-term vision includes delivering drugs or electrical stimulation through the same platform. The aim is a closed-loop system that not only senses problems but automatically treats them.

Funding came from a roster of federal and philanthropic backers: the National Institutes of Health, the National Science Foundation, the American Cancer Society, the Army Research Office, the U.S. Army Medical Research Acquisition Activity and the Heritage Medical Research Institute. Additional support flowed through Caltech’s Kavli Nanoscience Institute, which supplied clean-room space for fabricating the microfluidic channels.

The engineering behind the sensors is deliberately simple. Instead of using expensive optical components, the team relied on electrochemical detection, essentially measuring tiny currents that flow when NO or H₂O₂ reacts at a coated electrode. That choice keeps the bill of materials low and the form factor thin enough to wear inside a shoe or under clothing.

Yet simplicity on paper does not mean the device skirts sophistication. The machine-learning layer, trained on hundreds of hourly readings, adapts to each patient’s baseline and flags deviations that matter clinically.

For now, the bandage’s focus on two biomarkers is grounded in biology: both NO and H₂O₂ surge during the early inflammatory cascade and crash once healing is under way. By mapping those peaks and troughs, caregivers gain insight into whether tissue is progressing, plateauing or back-sliding toward infection.

Prototype units were fabricated at Caltech and shipped to USC for the trial. Participants wore the patch for varying intervals, during which standard care continued under physician supervision. Data streamed over Bluetooth to a clinical dashboard, where doctors compared algorithmic forecasts with their own bedside assessments.

Although the study was small, its prospective design and human focus set it apart from previous animal-only demonstrations. Gao’s team argues that showing feasibility in real-world clinics is a prerequisite for the regulatory approvals required to sell a medical-grade version.

Chronic wounds cost the health-care system billions annually and can lead to amputations or systemic infections. Tools that spot trouble early could help head off those expensive, life-altering outcomes. While telehealth cameras and consumer wearables can give a partial view, they cannot capture what is happening at the molecular level. That is where the iCare concept distinguishes itself.

If the next phase of trials confirms faster healing or fewer hospital visits, the bandage could move from prototype to prescription. Until then, the Caltech–USC team is already enrolling additional volunteers.