A functional outcome prediction model of acute traumatic spinal cord injury based on extreme gradient boost

This study aimed to create a model using a machine learning technique called Extreme Gradient Boost (XGBoost) to predict how well patients with acute spinal cord injuries (SCI) would recover one year after surgery. The model uses clinical data, MRI scans, and surgical timing to predict the Spinal Cord Independence Measure (SCIM) score, which indicates functional outcome. The XGBoost model was better at predicting outcomes than traditional linear models, and the patient's initial motor score and age were the most important factors in the prediction.

• 1
  A nonlinear regression prediction model was successfully constructed using XGBoost for patients with acute SCI.
• 2
  The nonlinear model predicted functional outcomes more accurately than traditional linear regression models.
• 3
  Baseline ASIA motor score (AMS) and age were the most important factors in predicting functional outcome.