Chronic wounds remain one of the most challenging complications of diabetes, affecting millions of people worldwide. Delayed healing, persistent inflammation and an increased risk of infection often lead to prolonged treatment, hospital admissions and, in severe cases, amputation. Conventional wound-closure methods such as sutures, staples and tissue adhesives can bring wound edges together, but they do not actively adapt to the body's healing process.
Researchers at Hanyang University in South Korea have now developed an intelligent microneedle patch that changes shape at body temperature to actively support wound closure while simultaneously delivering regenerative therapy and antibacterial protection. The technology combines artificial intelligence (AI), 4D printing, biomimicry, DNA nanotechnology and surface engineering into a single wound-healing platform. According to the researchers, the platform could pave the way for a new generation of responsive biomaterials that interact dynamically with biological tissues instead of functioning as passive medical devices.
Inspired by nature, optimized by AI
The design of the patch was inspired by Drosera capensis, a carnivorous sundew plant that captures insects through coordinated movement, adhesion and protective mechanisms. Mimicking these biological principles, the research team created shape-memory microneedles capable of bending after insertion into tissue.
The structures were produced using 4D printing, an emerging manufacturing technique that allows printed objects to change shape in response to environmental stimuli. In this case, the microneedles remain straight during application but automatically curve when exposed to physiological temperature (37°C), helping to draw wound edges together while maintaining stable contact with surrounding tissue. AI played a central role during development. Rather than relying on extensive trial-and-error experiments, the researchers employed machine-learning algorithms to predict how different material compositions and manufacturing conditions would influence shape recovery.
Among several AI models evaluated, Gaussian Process Regression proved most effective, accurately predicting material behaviour while also providing reliable estimates of uncertainty. This enabled the team to identify an optimal manufacturing window that balanced mechanical stability with rapid shape transformation. According to lead researcher Hyun-Do Jung, AI did more than optimise production. It translated biological inspiration into a predictable and programmable biomedical device with potential clinical relevance.
Regeneration and infection control
The smart patch offers more than mechanical wound closure. It incorporates adhesive DNA nanoparticles that gradually release regenerative molecules to stimulate tissue repair, while a zinc-treated surface provides antibacterial protection against common wound pathogens. Laboratory testing demonstrated that the microneedles rapidly adopted their programmed curved shape at body temperature, promoting stable tissue contact throughout the healing process. The platform also supported favourable responses from endothelial cells and fibroblasts, two cell types essential for blood vessel formation and tissue regeneration.
In addition, the zinc-coated surface showed strong antibacterial activity against both Escherichia coli and Staphylococcus aureus, two bacteria frequently associated with wound infections. Preclinical wound-healing experiments demonstrated faster wound closure and improved tissue regeneration compared with conventional wound-management approaches.
Beyond chronic wound care
Although further research and clinical validation will be required before the technology reaches patients, the researchers believe the underlying AI-guided 4D-printing strategy has applications well beyond chronic wound care. The same design principles could be used to develop soft biomedical robots, smart implants, tissue scaffolds and vascular stents capable of controlled shape transformation and long-term interaction with living tissue. Such devices could respond dynamically to physiological conditions while maintaining stable contact with biological structures.
For diabetic wound care in particular, the technology represents a shift from passive dressings toward intelligent biomaterials that actively participate in healing. By combining programmable materials with AI-driven design, the platform may ultimately help reduce complications, improve tissue regeneration and shorten recovery times for patients living with chronic wounds.
Microneedle patch technology
Earlier this year, researchers developed a 3D-printed microneedle patch that significantly improves the delivery of live virus vaccines. The study introduced a “pillar-guided” design in which a 3D-printed backing layer directs vaccine material into the tips of dissolvable microneedles. This shortens drying time, preserves more viable virus during manufacturing and increases the effective vaccine dose. In preclinical studies, mice vaccinated with the redesigned patches developed stronger virus-specific immune responses and were protected against lethal SARS-CoV-2 infection.
Microneedle array patches offer several advantages over conventional injections: they are minimally invasive, can be self-administered and remain stable at room temperature, reducing reliance on cold-chain logistics. The researchers believe the technology could improve vaccine access in low-resource settings and be adapted for other viral vaccines, supporting future pandemic preparedness and routine immunization programs.
