Smarter Wound Monitoring With AI

A new medical device created at the University at Buffalo has the potential to save time for health care providers while also safeguarding patients.

Using radar and artificial intelligence, the system can see through dressings to assess wounds and other skin conditions, eliminating the need to remove and reapply bandages—a time-consuming process that exposes these vulnerable areas to germs and other pathogens.

“Our system has great potential to reduce secondary infections, cut costs associated with caring for those infections, and save nurses and other health care providers time so they can concentrate on other duties,” said Wenyao Xu, professor of computer science and engineering at UB and corresponding author of a study published in IEEE Internet of Things Journal describing the new technology.

The device utilizes millimeter radio waves—the same ones used by radar guns, airport security scanners and 5G cell phone networks. With the new system, these waves measure the moisture content of a wound, which is a critical indicator of overall wound health.

To ensure the device filters out obstructions such as walls and other medical instruments that can interfere with the radio waves, the team created a complex algorithm that focuses only on the signal from the wound area. The researchers also developed and trained an artificial intelligence model that interprets the signals and converts them into accurate wound moisture data.

After building a prototype for initial tests and creating 60 “wound phantoms” of differing shapes, sizes and moisture levels, the researchers found the system was 99.45% accurate in detecting moisture levels and was able to see through two layers of gauze—a thickness commonly used in wound care. On human skin with fake wounds, the system was roughly 95.5% accurate.

Also impressive, Xu said, is that the system worked well for different skin tones, ages and genders, which suggests its reliability for a range of patients.

The researchers are currently testing the system, in partnership with UBMD Surgery, on real and complex human wounds. The goal of these clinical trials is to eventually reduce the wound scanning time from a minute or more to under 10 seconds, make the images clearer, and bring the technology to clinical applications in wound care and healing monitoring.