Machine Learning Models for Predicting Return-to-Play Timelines After Sports-Related Concussion

This study developed machine learning models using data from 971 college athletes to predict recovery timelines and recovery type (typical vs. prolonged) following sports-related concussion. Post-concussion symptoms and cognitive performance measured 24-48 hours post-injury were the strongest predictors, with classification models achieving 89% accuracy in distinguishing typical from prolonged recovery trajectories.