An Efficient and User-Friendly Hybrid Approach for Classifying and Predicting Cardiovascular Disease Risks

For any disease, health symptoms indicate its status in terms of severity whether medium, low, or high and their impact on life survival. There are stages of the disease that need to be classified. Among all diseases, severity is particularly significant in the case of heart disease. While health symptoms often suggest the presence of a disease, accurately predicting severity and categorizing risks, especially for complex conditions like cardiovascular disease (CVD), remains crucial. There are numerous factors that indicate the risk for cardiovascular problems. These include not only traditional risk factors like age and family history, but also genomic data, coronary artery scores, and carotid intima-media thickness analyzed through imaging studies, along with insights from biochemical inflammatory markers. Among these, warning signs such as shortness of breath, discomfort, fatigue, irregular heartbeat, leg numbness, and chest pain can affect the risk of a heart attack. The risks can be classified as modifiable or non-modifiable, time-frame risks, and low or severe risks. To predict and classify the risk in CVD, the integration of machine learning models and rule development can help determine the classification of risks. The hybrid approach is designed to identify and classify the risks into appropriate categories. These risks may affect health based on their severity. The computational process involved must be measured against accuracy and efficiency. The proposed hybrid model is expected to demonstrate exceptional performance compared to existing approaches, and these measures will be represented in graphical aids for better understanding and to ensure quality. This study explores a hybrid approach leveraging both machine learning and rule-based systems to achieve the objectives of the proposal.