Abstract
Artificial intelligence (AI) is revolutionizing regenerative medicine, particularly in advancing mesenchymal stem cell (MSC) therapies and smart biomaterials. This review highlights AI's role in two core areas: First, at the biological level, AI can be used to predict MSC differentiation, immunomodulatory function, and therapeutic potential by analyzing multi-omics and imaging data, deciphering heterogeneity and improving precision. For instance, deep learning models based on MSCs' morphology can successfully predict the differentiation propensity and uncover the regulatory networks underlying the intrinsic heterogeneity. Second, in engineering, AI shifts material design from trial-and-error to data-driven approaches. Machine learning models correlate material parameters with biological properties, enabling optimized screening. Furthermore, generative AI can be used to tailor novel materials through inverse design to achieve targeted properties like accelerated wound healing. However, the current development in this field remains constrained by several severe challenges, including the fragmented nature of existing research evidence, the insufficient reproducibility of model predictions in independent cohorts, and the significant translational gap from computational predictions to in vivo validation. Future research must not only demonstrate potential but also urgently address these fundamental and translational bottlenecks.