A Smart Recommender System to Stratify Heart Attack Risk

Abstract

Heart diseases are a major cause of mortality and morbidity. Faster detection of life-threatening emergency events and an earlier start of the therapy would save many lives and reduce successive disabilities. We developed an automated smart recommender system (Explainable Artificial Intelligence) for heart attack prediction and risk stratification. This study is focused on an improved recommender approach to identify key risk factors impacting risk stratification for a particular patient. To help patients and clinicians better understand specific risk factors associated with heart attack and the degree of association, we used CatBoost classifier to build the model and SHAP to interpret its results. The risk factors that can be measured using smart monitoring wearable devices are called dynamic, while other risk factors obtained from users directly are static. The system collects both factor types to predict a heart attack risk. We created a Django-based online application that uses patient data to update medical information about an individual’s heart attack risk. Moreover, the individual can quickly locate nearby doctors and hospitals from the application. The smart recommender system achieved high accuracy in predicting a patient risk level with an average AUC of 0.85. Utilizing the SHAP inter pretation technique, we provided insights into the reasoning behind the predictions, including group-based and patient-specific explanations. Additionally, we employed a smart monitoring wearable device, such as a Smartwatch to automatically gather dynamic risk factors from the patient. The recommender system is cost-effective, easy to use, and portable since the main component of the system is commonly available on smartwatches, smartphones, and the Django framework. Early detection can improve patient management and lower heart attack risk while timely therapy aids in avoiding subsequent disabilities.