Using machine learning to predict stroke-associated pneumonia in Chinese acute ischaemic stroke patients
1School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, Nanjing, China.
European Journal of Neurology
|May 7, 2020
Summary
Machine learning accurately predicts stroke-associated pneumonia in Chinese patients with acute ischemic stroke. The XGBoost model offers a superior tool for early risk identification, improving patient outcomes.
Area of Science:
- Neurology
- Medical Informatics
- Public Health
Background:
- Stroke-associated pneumonia (SAP) is a frequent and serious complication following acute ischemic stroke (AIS).
- Early identification of high-risk patients is crucial for preventing SAP.
- Existing prediction models have limited clinical application.
Purpose of the Study:
- To develop and validate a machine learning (ML) model for predicting SAP in Chinese AIS patients.
- To compare the performance of ML models against established clinical scores.
Main Methods:
- A prospective cohort of 3160 AIS patients was used, randomly divided into training and testing sets.
- Five ML models were developed: logistic regression, support vector machine, random forest, XGBoost, and deep neural network.
- Model performance was evaluated using the area under the receiver operating characteristic curve and compared with ISAN and PNA scores.
Main Results:
- The XGBoost model demonstrated the highest predictive performance with an area under the curve of 0.841 (sensitivity 81.0%, specificity 73.3%).
- The XGBoost model significantly outperformed the ISAN and PNA scores in predicting SAP.
- The best-performing model utilized six common variables.
Conclusions:
- The XGBoost model provides an optimal prediction tool for SAP in Chinese AIS patients.
- This ML-based approach surpasses traditional scoring systems like ISAN and PNA.
- The developed model can aid in early risk stratification and prevention strategies for SAP.
