Deep Learning in Spine Surgery

Abstract Deep learning is increasingly impactful for healthcare research and delivery. In orthopaedics, there has been a significant increase in the number of publications using deep learning for interpreting radiographs; however, there are few applications of deep learning specific to spine surgery. In this review, we discuss the potential of deep learning within spine surgery in four contexts: diagnosis (1), prognosis (2), patient management and integration with virtual/augmented reality (3), robotic surgery and biomedical wearables (4). Additionally, we discuss current literature, future potentials and provide takeaways for clinicians who wish to apply deep learning in their research and practice.