Artificial Intelligence in Botulinum Toxin Injections: A Mini-Review of Current Applications, Challenges, and Translational Perspectives.

Botulinum toxin (BoNT) injection is widely used in aesthetic dermatology, plastic surgery, neurology, and rehabilitation, but treatment still depends on the clinician's experience, anatomical judgment, and iterative dose adjustment. Artificial intelligence (AI) may support BoNT practice through facial analysis, anatomical mapping, treatment simulation, response prediction, image-guided injection, and objective outcome assessment. Current evidence includes chatbot-assisted planning, deep learning analysis of facial expression, magnetic resonance imaging-based prediction of dystonia response, multimodal machine learning for spasticity, computational diffusion modeling, and AI-assisted ultrasound interpretation. This mini review, based on PubMed-indexed and related peer-reviewed literature, summarizes representative AI applications across the BoNT treatment pathway and emphasizes a cautious clinical framework in which AI supports, but does not replace, physician judgment. Evidence remains preliminary and heterogeneous, with limitations related to small sample sizes, retrospective designs, and limited external validation. However, this mini-review synthesizes the most recent and relevant high-quality studies and provides a timely, structured overview of emerging AI applications to support clinical decision-making in BoNT practice. Future translation should prioritize prospective validation, formulation-aware dose modeling, transparent governance, and physician-supervised implementation.