Periodontal disease is a prevalent chronic inflammatory condition, and its heterogeneity and complex pathophysiology pose significant diagnostic challenges. Traditional methods remain limited in precision and early detection. With the advancement of precision medicine, data-driven diagnostics are shifting periodontal assessment from a “one-size-fits-all” model to a more precise approach. Moreover, numerous studies highlight the vast potential of artificial intelligence (AI) in periodontal risk assessment and diagnosis. Integrating AI with data-driven diagnostic models enables deeper analysis, potentially surpassing the limitations of conventional empirical medicine and establishing a new paradigm for precision oral healthcare. This review explores recent advances in data-driven and AI-based strategies for periodontal disease diagnosis, emphasizing the multidimensional integration of biological features analysis and providing new insights into periodontal precision diagnostics.

Traumatic brain injury (TBI) contributes to the key causative elements of neurological deficits. However, no effective therapeutics have been developed yet. In our previous work, extracellular vesicles (EVs) secreted by stem cells from human exfoliated deciduous teeth (SHED) offered new insights as potential strategies for functional recovery of TBI. The current study aims to elucidate the mechanism of action, providing novel therapeutic targets for future clinical interventions. With the miRNA array performed and Real-time PCR validated, we revealed the crucial function of miR-330-5p transferred by SHED-derived EVs (SHED-EVs) in regulating microglia, the critical immune modulator in central nervous system. MiR-330-5p targeted Ehmt2 and mediated the transcription of CXCL14 to promote M2 microglia polarization and inhibit M1 polarization. Identified in our in vivo data, SHED-EVs and their effector miR-330-5p alleviated the secretion of inflammatory cytokines and resumed the motor functional recovery of TBI rats. In summary, by transferring miR-330-5p, SHED-EVs favored anti-inflammatory microglia polarization through Ehmt2 mediated CXCL14 transcription in treating traumatic brain injury.