PREDICTING THE RISK OF HEART DISEASE IN DIABETIC PATIENTS USING MACHINE LEARNING TECHNIQUES

Heart disease is a leading cause of death worldwide, with diabetic individuals facing an especially high risk due to complex metabolic and cardiovascular disruptions linked to diabetes. Early and accurate identification of high-risk diabetic patients is essential for timely intervention and prevention. However, traditional cardiovascular risk scoring systems often underperform in diabetic populations because they are developed from general population data and do not fully account for diabetes-specific factors. To address, this work was undertaken to develop and validate a machine learning-based predictive model tailored specifically for assessing heart disease risk in type 2 diabetic patients. Using a retrospective dataset of 1,000 adult diabetic patients, which included demographic, clinical, and laboratory variables, the data was preprocessed to handle missing values, encode categorical data, and scale features. Feature selection techniques identified key predictors such as HbA1c levels, age, systolic blood pressure, and cholesterol. Several machine learning algorithms were evaluated—including K Nearest Neighbors, Support Vector Classifier, Decision Tree, and Random Forest— with the Random Forest Classifier achieving the best performance: 89% accuracy and an AUC-ROC of 0.92. This highlights the importance of integrating diabetes-specific markers with traditional cardiovascular risk factors. The final model was deployed as an accessible web-based application, enabling healthcare providers and patients to input relevant health data and receive personalized heart disease risk assessments. This tool supports data-driven clinical decisions, aiming to enhance early detection and targeted management of cardiovascular risks in diabetic individuals, ultimately improving patient outcomes.