Re: Global research trends in AI-related foot and ankle surgery research: A bibliometric and visualized study

Dear Editor,

We read with interest the recent article by Zhang and Du, Global research trends in AI-related foot and ankle surgery research: A bibliometric and visualized study. ¹ The authors should be commended for offering the first quantitative overview of this rapidly expanding domain. Their analysis illuminates global leadership patterns, the steady growth of publications, and the emergence of diabetic foot ulcers and computer-aided diagnosis as critical hotspots.

While valuable, we believe three additional dimensions merit emphasis to strengthen the translational relevance of this field.

First, methodological limitations of bibliometrics. The reliance on Web of Science data may underrepresent high-impact clinical studies indexed primarily in PubMed or Scopus, particularly case-control or prospective orthopaedic trials. Citation-based mapping also risks underweighting cutting-edge but recently published studies, such as AI-assisted ankle fracture detection validated in emergency departments within the past 2 years. ² A more comprehensive approach integrating multiple databases and normalizing for citation lag would better align bibliometric signals with real-world innovation.

Second, clinical pathways beyond diagnostic imaging. Current bibliometric clusters heavily emphasize diabetic foot ulcers and fracture detection, but little attention is given to perioperative decision support in arthroscopy, tendon reconstruction, or total ankle arthroplasty. For example, machine learning has been applied to predict postoperative venous thromboembolism after ankle fracture ³ and to optimize rehabilitation trajectories following Achilles tendon repair. ⁴ These clinically grounded pathways highlight AI’s potential impact on surgical planning, implant longevity, and patient-specific functional recovery—areas insufficiently captured in bibliometric visualization alone.

Third, emerging paradigms in AI deployment. While the article notes challenges of interpretability, it does not explore new approaches such as federated learning for privacy-preserving collaboration across institutions, ⁵ or multimodal models that integrate imaging, kinematic, and electronic health record data. These paradigms may overcome current bottlenecks of data heterogeneity and limited external validation, providing more robust and equitable AI adoption across diverse health systems.

In summary, Zhang and Du provide an important foundation for understanding the trajectory of AI in foot and ankle surgery. To move beyond descriptive bibliometrics growth, future work should triangulate multiple data sources, critically link trends to operative and postoperative care, and explore emerging paradigms such as federated and multimodal AI. Only then will bibliometrics not merely describe the landscape but also anticipate transformative directions in orthopaedic surgery.