Chronic inflammatory pathologies, such as diabetic foot ulcers (DFU), are predominantly governed by persistent oxidative stress and impaired mechanisms of tissue regeneration.1 Recently, plant-derived extracellular nanovesicles (PDEVs) have gained the spotlight as natural nanocarriers with the highly biocompatible manner and intrinsic therapeutic effects. These vesicles typically include various antioxidants, anti-inflammatory and bioactive molecules such as polyphenols, flavonoids, proteins, and membrane structures that facilitate cellular internalization. This study proposes a new therapeutic platform for chronic wounds, highlighting their capacity to regulate reactive oxygen species (ROS), augment angiogenesis, and promote tissue repair within diabetic microenvironments.

Coating is one effective approach to control the fast corrosion of magnesium. Micro arc oxidation (MAO) coatings have shown significant results in slowing down the degradation process. Although experiments have proven the efficacy of this method in studying corrosion, there remains a need to identify more time and cost-effective solutions. Finite element analysis offers a potential solution, but the complexity of the factors involved in the MAO process makes machine learning methods a more efficient and time-saving alternative for predicting corrosion behavior. This study aims to expand existing machine learning models from predicting corrosion in uncoated magnesium to innovatively forecasting the degradation of MAO-coated samples, focusing on predicting corrosion current density and corrosion potential based on electrochemical corrosion test data. The results showed that the use of a log prediction method demonstrates a significant ability to predict the log of the corrosion current density, as the exact numerical value can vary between individuals. Additionally, the log prediction method effectively predicts corrosion potential with a minimal mean absolute error (MAE).