Decision Tree Analysis for Predicting Recovery in Patients with Sudden Sensorineural Hearing Loss

There is growing interest in predicting treatment outcomes, aiding in risk classification and identifying significant factors in patient management. Sudden Sensorineural Hearing Loss (SSNHL) is a condition that requires immediate medical attention, and timely recognition and treatment can enhance patient hearing. This study employed decision tree modeling to forecast recovery outcomes in SSNHL patients treated at Ghaem Hospital in Mashhad. Data from ۳۳۰ patients were analyzed, revealing an ۸۴.۵% recovery rate. The CART algorithm was utilized, identifying ۱۱ variables, with significant predictors including age, time to treatment initiation, and gender. The model demonstrated a mean residual error of ۰.۴۴ and a classification error rate of ۹%. Findings suggest that decision trees can effectively predict treatment outcomes and important factors for SSNHL.

کلیدواژه ها:

Decision Tree Modeling ، Treatment Outcomes ، CART Algorithm