Artificial Intelligence-Based Ensemble Learning Model for Prediction of Hepatitis C Disease.

Area of Science:

• Hepatology
• Medical Informatics
• Artificial Intelligence

Background:

• Hepatitis C poses a significant global health risk, potentially leading to severe liver damage and mortality.
• Early detection and intervention are crucial for effective treatment, which can cure most patients within 8-12 weeks.
• Preventing transmission, particularly among injecting drug users, remains a key public health challenge.

Purpose of the Study:

• To develop and evaluate an AI-based ensemble model for predicting Hepatitis C disease.
• To assess the model's capability in predicting advanced fibrosis by integrating clinical data and blood biomarkers.
• To compare the accuracy of the ensemble model against individual machine learning algorithms for Hepatitis C prediction.

Main Methods:

• Utilized three distinct machine learning approaches for training classification models.
• Developed an AI-based ensemble model by integrating multiple algorithms.
• Validated the ensemble model using a dedicated dataset, incorporating clinical data and blood biomarkers.

Main Results:

• The AI-based ensemble model achieved a higher accuracy of 95.59% in predicting Hepatitis C.
• Individual algorithms showed strong performance, with Multi-layer Perceptron (MLP) at 94.1% and Bayesian Network at 94.47%.
• The ensemble approach demonstrated superior predictive accuracy compared to standalone machine learning techniques.

Conclusions:

• AI-based ensemble learning models offer a more precise and accurate method for predicting Hepatitis C disease.
• This advanced prediction capability can aid in early diagnosis, risk mitigation, and improved treatment outcomes.
• The study highlights the potential of ensemble models in enhancing the accuracy of diagnosing Hepatitis C.