Abstract: Postoperative pulmonary complications are prevalent after lung cancer surgery, severely impacting patient recovery. Traditional pulmonary rehabilitation faces challenges such as lack of out-of-hospital supervision, sharp decline in adherence, and insufficient personalization. AI-driven respiratory training provides a new pathway to address these issues. This review summarizes the application of AI-driven respiratory training systems in perioperative rehabilitation management for lung cancer patients. It highlights its value in personalized assessment and regimen design, real-time monitoring and feedback. Furthermore, the review analyzes the transformation of the nurse's role from a "reactive executor" to a "data-driven decision-maker" and proposes multi-dimensional strategies to address challenges in legal liability, data governance, and policy support, aiming to provide references for the development of "smart nursing" models.

Key words: Artificial Intelligence; smart nursing; lung cancer; perioperative period; pulmonary rehabilitation; respiratory training