OCT Powered by AI-Based Analytics Gives a Glimpse into Wound Healing

No matter the size or severity, wounds on human skin are difficult to monitor while they heal. Biopsies disrupt the wound site and are too invasive for routine, repeated monitoring, and most medical imaging devices that could do the job are large, expensive and booked up with more pressing diagnostics. Clinicians typically resort to visual inspection or quick measurements of the wound’s size over time.

Based on research completed as part of a multi-year collaboration with Nokia Bell Labs, biomedical engineers at Duke University are developing a solution. Using a custom-built optical coherence tomography (OCT) imaging system together with artificial intelligence (AI) models grounded in a deep understanding of tissue regeneration, researchers have shown they can accurately and objectively measure the progress of wounds healing over time.

Using their new approach, the researchers also show that a hydrogel under development to improve wound healing works better with stiffer mechanical properties. The results are a two-for-one boon in a challenging area for both clinicians and researchers.

The research appears online March 20 in the journal Cell Biomaterials.

“Wound healing is a complex process, and what we see on the surface doesn’t always reflect what’s happening underneath,” said Sharon Gerecht, chair and the Paul M. Gross Distinguished Professor of Biomedical Engineering at Duke. “For more than a decade, my lab has developed hydrogel-based therapies to guide tissue healing and regeneration. Partnering with Nokia Bell Labs allowed us to combine advanced optical imaging and AI has given us unprecedented insights into how biomaterials induce healing beneath the surface.”

OCT is best known for its role in eye care, where it provides 3D images of the back of the eye to help diagnose and monitor retinal diseases. Now, researchers have adapted that same depth-resolved imaging capability to wound healing, using light to non-invasively visualize tissue architecture and blood flow beneath the skin.

Turning those rich scans into meaningful biological insights, however, requires more than imaging alone. Parsing through the information demands quantitative tools that can rapidly interpret large volumes of complex data. That is where the collaboration with Nokia Bell Labs proved essential.

Researchers at Nokia Bell Labs over the multi-year project, developed a custom OCT system along with AI-driven analytical methods that were trained on imaging datasets acquired in the Gerecht lab. This OCT-AI platform enabled the team to move beyond simple visualization, making it possible to automatically quantify how tissue structure and vascular dynamics evolve over time as well as objectively assess the degree of healing.

To evaluate the technology, the collaborative team applied it to wounds in mice treated with a hydrogel developed in the Gerecht lab. To demonstrate the broader research potential of this platform, they compared hydrogels with relatively soft and relatively stiff mechanical properties.