AI meets medicine with “a-Heal,” a breakthrough smart bandage that senses, learns, and heals wounds up to 25% faster.
- The a-Heal device uses AI and bioelectronics to accelerate wound healing by about 25% in preclinical studies
- Its AI physician analyzes images, determines healing stages, and adjusts medication or electric fields automatically
- Reinforcement learning and Deep Mapper technology allow the system to adapt to each patient’s healing process in real time
A wound heals in various stages, including clotting to prevent bleeding, immune system reaction, scabbing, and scarring. A wearable device called “a-Heal”, created by engineers at the University of California, Santa Cruz, tries to improve each stage of the procedure (1). The technology detects the state of healing with a tiny camera and AI and then delivers a treatment in the form of medication or an electric field. The system responds to the patient’s individual wound healing process by providing personalized treatment.
Novel Device Accelerates Wound Healing
The portable, wireless gadget may make wound care more accessible to patients in distant places or with limited mobility. Initial preclinical data, published in the journal npj Biomedical Innovations, reveal that the device effectively accelerates the healing process.
Device with Camera, Bioelectronics, and AI Speeds Up Wound Healing
A team of UC Santa Cruz and UC Davis researchers, sponsored by the DARPA-BETR program and led by UC Santa Cruz Baskin Engineering Endowed Chair and Professor of Electrical and Computer Engineering (ECE) Marco Rolandi, created a device that combines a camera, bioelectronics, and AI to speed up wound healing. According to the researchers, the integration in one device creates a “closed-loop system”–one of the first of its kind for wound healing.
The device employs an inbuilt camera, developed by fellow Associate Professor of ECE Mircea Teodorescu and reported in a Communications Biology article, to photograph the wound every two hours. The pictures are loaded into a machine learning (ML) model developed by Associate Professor of Applied Mathematics Marcella Gomez, which the researchers refer to as the “AI physician” and runs on a nearby computer.
“It’s essentially a microscope in a bandage,” Teodorescu explained. “Individual images say little, but over time, continuous imaging lets AI spot trends, wound healing stages, flag issues, and suggest treatments.”
AI Physician Determines Wound Stage and Administers Treatment
The AI physician analyzes the image to determine the wound stage and compares it to where the wound should be on a timeline for optimal wound healing. If the image shows a lag, the ML model administers a treatment: either medicine given by bioelectronics or an electric field, which can stimulate cell migration toward wound closure.
The device delivers fluoxetine topically, a selective serotonin reuptake inhibitor that regulates serotonin levels in the wound and promotes healing by reducing inflammation and improving wound tissue closure. The dose, determined by preclinical investigations by the Isseroff lab at UC Davis to enhance healing, is delivered using bioelectronic actuators on the Rolandi-developed device. The device also delivers an electromagnetic field that has been tuned to aid healing and was developed previously by UC Davis’ Min Zhao and Roslyn Rivkah Isseroff.
The AI physician decides the appropriate drug dosage and the amount of the administered electric field. After the therapy has been applied for a set amount of time, the camera takes another image, and the procedure begins again.
Human Intervention to Fine-Tune Treatment
While in operation, the device sends photos and data, such as healing rate, to a secure web interface, allowing a human physician to actively intervene and fine-tune treatment as necessary. For ease of usage and security, the device attaches directly to a commercially available bandage.
To examine the potential for clinical application, the UC Davis team tested the device in preclinical wound models. In these investigations, wounds treated with a-Heal healed around 25% faster than normal treatment. These findings emphasize the technology’s potential not just for expediting acute wound closure, but also for restarting halted healing in chronic wounds.
Reinforcement Learning Powers a Smarter Wound-Healing System
The AI model utilized for this system, directed by Assistant Professor of Applied Mathematics Marcella Gomez, mimics the diagnostic strategy employed by clinicians through a reinforcement learning approach outlined in a study published in the journal Bioengineering.
Reinforcement learning is a technique in which a model is created to attain a specified end goal, and it learns how to do so through trial and error. In this case, the model is given the aim of reducing time to wound closure and is rewarded for making progress toward it. It continuously learns from the patient and adjusts its treatment strategy.
The reinforcement learning model is driven by Deep Mapper, an algorithm developed by Gomez and her students and published in a preprint research, which processes wound images to quantify the stage of healing in relation to typical progression and maps it along the healing trajectory.
As time passes with the device on a wound, it develops a linear dynamic model of previous healing and uses it to predict how the healing process will proceed.
“Having the image isn’t enough; you also need to process it and put it in perspective. “Then you can use the feedback control,” Gomez explained.
This technique enables the algorithm to learn in real time the effect of the drug or electric field on healing, and it leads the reinforcement learning model’s iterative decision-making on how to modify the drug concentration or electric-field intensity.
The research team is currently investigating the device’s potential to aid the healing of chronic and infected wounds.
Frequently Asked Questions
What is the a-Heal device?
It’s a wearable AI-powered smart bandage that monitors wounds and delivers precise treatments automatically.
How does AI help in wound healing?
The AI model analyzes wound images, tracks healing progress, and adjusts therapy through reinforcement learning.
Is this technology safe for clinical use?
It has shown strong preclinical success and is now being studied for use in chronic and infected wounds.
References:
- Towards adaptive bioelectronic wound therapy with integrated real-time diagnostics and machine learning–driven closed-loop control
(Li, H., et al. (2025). Towards adaptive bioelectronic wound therapy with integrated real-time diagnostics and machine learning–driven closed-loop control. npj Biomedical Innovations. doi.org/10.1038/s44385-025-00038-6)
Source-Medindia