This article discusses the development of a predictive model for predicting spinal cord injury (SCI) in patients with spinal tuberculosis (STB) using machine learning algorithms and clinical data from multiple hospitals. The model was validated and tested on a prospective cohort of STB patients.