Machine learning model identifies patient gait speed throughout the episode of care, generating notifications for clinician evaluation

Utilization of ML to predict gait recovery following total joint replacement is feasible and provides results with excellent specificity. This model will allow inclusion of additional data for retraining as patient populations evolve. Clinician feedback regarding notifications, including resulting actions and outcomes, can be used to further inform the model and improve clinical utility.