AI-Based Heart Disease Diagnosis and Risk Assessment

Cardiovascular diseases (CVDs) remain a leading cause of mortality worldwide, emphasizing the necessity of advanced, efficient, and accessible diagnostic methods. This study explores the application of Artificial Intelligence (AI) and Machine Learning (ML) techniques to enhance heart disease diagnosis and risk assessment, particularly in resource-limited settings such as Libya. By leveraging ensemble learning methods, including eXtreme Gradient Boosting (XGBoost) and Multilayer Perceptron (MLP) Neural Networks, we improve predictive accuracy and enable early detection. The Cleveland Heart Disease Dataset was utilized to train and evaluate these models in both multi-class and binary classification tasks. Experimental results indicate that XGBoost, specifically with DMatrix implementation, achieved the highest accuracy (73.33%) in multi-class classification, while both XGBoost and MLP demonstrated significant improvements in binary classification, reaching an accuracy of 91.67%. These findings highlight the potential of AI-driven diagnostic systems to address the challenges of delayed diagnosis and limited medical resources in Libya and similar low-resource environments. The proposed AI-based system provides a user-friendly and efficient tool to assist healthcare professionals in clinical decision-making, improve patient outcomes, and facilitate the integration of AI into global public health infrastructures.