Development and Validation of Artificial Intelligence - Based Cardiovascular Disease (AI-CVD) Risk Score

Abstract

Introduction: Cardiovascular diseases are one of the most prevalent diseases in India amounting for nearly 30% of total deaths. A study on a large-scale patient data is needed to determine, analyze and interpret risk factors associated with Cardiovascular Diseases (CVDs) in Indian population. Objective: Objective of this study is to determine the risk factors associated with cardiovascular diseases and develop a CVD risk score model to predict a cardiovascular event in the next 5 - 7 years. Method: The study included data corresponding to 31,599 participants aged between 18 to 91 from 2010 to 2017 from Apollo Hospitals centres located in India. A multi-step risk factors selection process was used to determine the clinical and life style related risk factors. Cox-proportional hazard model trained on the studied population yielded Area under ROC (Receiver Operating Characteristic) curve (AUC-ROC) score of 0.830 on 5-fold cross validation, while deep survival model yielded a 0.853 AUC-ROC score. Results: The study shows higher significance for risk factors like diabetes, hypertension, dyslipidemia, diastolic blood pressure and chewing tobacco in Indian population. Conclusion: This is a novel cardiac risk related study on Indian population using retrospective data with expanded set of risk factors and use of artificial intelligence. Trial Registration: Registered with Clinical Trial Registry of India (CTRI) - CTRI/2019/07/020471 Funding Statement: There were no available or separate funding for this research. Declaration of Interests: All authors declare that there are no competing/conflicting interests. Ethics Approval Statement: The prospective research has been approved in all of nine center’s Institutional Ethics Committees and had been deliberated with over 100 Cardiologists nationally and internationally.