Prediction of Prognosis in Emergency Trauma Patients with Optimal Limit Gradient Based on Grid Search Optimal Parameters

This study proposes an innovative use of optimized extreme gradient enhancement based on grid search best parameters (Extreme Gradient Boosting, XGBoost) machine learning analyzer evaluates the prognostic state of emergency patients with severe trauma, with two prognostic outcomes, including rehabilitation and death. For patients with a trauma severity score (Injury Severity Score, ISS) greater than 15 in the ISS category of “serious trauma”, the use of XGBoost collects 3 parts of data into a database through emergency trauma patients to the prognostic result analysis, respectively, for hospitalization injury data, physician evaluation data, physician disposal data, and analysis prognotation status. After experimenting with the results of different methods, data attributes and predictive prognosis systems use the XGBoost method with a higher accuracy rate, thus achieving the goal of cost studies, so that medical personnel can learn about some vital signs and trauma scores and other data, you can estimate the prognostic status of patients, better treat emergency trauma patients, improve the probability of prognosis rehabilitation and reduce the risk of fatal.