AI-Guided Microneedles Mimic Carnivorous Plants to Heal Diabetic Wounds

Associate Professor Hyun-Do Jung and his team developed the platform to overcome the limitations of passive wound-closure methods like sutures or staples. The microneedles respond to the 37°C environment of the human body by bending into a specific shape, ensuring sustained contact with the wound site. This mechanical action is paired with a zinc-treated surface to combat bacteria and adhesive DNA nanoparticles that foster the growth of fibroblasts and endothelial cells.

To perfect the material's response, the researchers utilized Gaussian Process Regression to optimize the 4D-printing process. This machine-learning approach allowed the team to predict how different compositions would behave, significantly cutting down on trial-and-error laboratory cycles. The resulting device demonstrated effective antibacterial activity against Escherichia coli and Staphylococcus aureus during preclinical testing, while accelerating the overall closure of treated wounds. Published in Advanced Materials, the research suggests that this synthesis of AI and biomimicry could eventually extend to soft robotics and smart implants that require precise, programmable motion within biological environments.