Enhancing Early Detection of Neonatal Hearing Loss: A Machine Learning Risk Stratification Approach

Highlight

• Development of a machine learning-based risk stratification tool to predict hearing loss in high-risk neonates using clinical parameters.
• The XGBoost model demonstrated superior predictive performance with 85.2% accuracy and an area under the curve (AUC) of 87.1%.
• NICU stay duration and family history emerged as the most influential risk factors via SHAP interpretability analysis.
• A web-based clinical decision support application has been implemented for real-time risk assessment to assist clinicians in early identification and intervention.

Study Background