Improving Hepatitis B outcome prediction with ensemble machine learning: A study on predictive models and...

Area of Science:

• Medical Informatics
• Computational Biology
• Public Health

Background:

• Hepatitis B virus (HBV) infection poses a major global health challenge, leading to severe liver conditions like cirrhosis and liver cancer.
• Early diagnosis and intervention are crucial, particularly in resource-limited areas, to reduce the impact of HBV.
• Predictive modeling can enhance patient management and outcomes for HBV infection.

Purpose of the Study:

• To evaluate the efficacy of diverse machine learning (ML) techniques for predicting patient outcomes in Hepatitis B virus (HBV) infection.
• To identify key risk factors associated with mortality in HBV patients.
• To enhance the interpretability of ML models used in HBV outcome prediction.

Main Methods:

• Feature selection was performed using the Chi-squared test.
• Class imbalance was addressed using the Synthetic Minority Over-sampling Technique (SMOTE).
• Machine learning models including Support Vector Machine (SVM), Logistic Regression (LR), and a Voting Classifier were trained and evaluated. Model interpretability was enhanced using SHAP and LIME.

Main Results:

• Individual models, Support Vector Machine (SVM) and Logistic Regression (LR), achieved 92.5% accuracy.
• A Voting Classifier combining SVM and LR improved prediction accuracy to 95%.
• Elevated levels of certain risk factors, particularly in older individuals, were significantly associated with an increased risk of mortality.

Conclusions:

• Machine learning models demonstrate high accuracy in predicting HBV patient outcomes.
• The findings highlight the importance of specific risk factors and patient age in determining prognosis.
• These predictive insights can inform clinical decision-making and public health strategies for managing HBV infection.