TITLE:
Statistical Analysis of Cardiovascular Diseases Dataset of BRFSS
AUTHORS:
Sushant Kumar Gupta, Ashank Anshuman, Aakarshit Uppal, Indrajit Mukherjee
KEYWORDS:
Deep Learning, Predictive Models, Bioinformatics, Healthcare, Medicine
JOURNAL NAME:
Open Access Library Journal,
Vol.11 No.10,
October
31,
2024
ABSTRACT: Cardiovascular Diseases (CVDs) remain a leading cause of death in the United States. These diseases, including coronary heart disease, heart attack, and stroke, pose significant health risks. Accurate prediction of CVD probability can aid in prevention and management. To address this challenge, we analyzed data from the Behavioral Risk Factor Surveillance System (BRFSS) spanning 1995-2017. We developed innovative methods to handle missing data and normalize values. Deep learning models were employed to predict risk factors and, subsequently, the likelihood of CVDs. Our models were implemented using TensorFlow and trained on a high-performance computing server. The models accurately predicted risk factors with over 90% accuracy, enabling targeted interventions. We successfully predicted CVD probability with greater than 95% accuracy, providing valuable insights for healthcare providers. An online portal was developed to forecast CVD trends over the next 31 years, facilitating proactive planning and resource allocation.