Comparison of Data Mining Algorithms in Prediction of Coronary Artery Diseases Using Yazd Health Study (YaHS) Data

Azam Barzegari; Seyede Fatemah Noorani; Masoud Mirzaei

doi:10.18502/ssu.v31i7.13693

Azam Barzegari Shahid Sadoughi University of Medical Sciences, Yazd, Iran.
Seyede Fatemah Noorani Faculty of Engineering, Department of Computer Engineering and Information Technology, Payame Noor University, Tehran, Iran.
Masoud Mirzaei Cardiovascular Research Center, Shahid Sadoughi University of Medical Sciences, Yazd, Iran.

DOI: https://doi.org/10.18502/ssu.v31i7.13693

Keywords: Data mining, Screening, Coronary heart Disease, Naive Bayes, Random Forest, YaHS.

Abstract

Introduction: Cardiovascular diseases, including ischemic heart disease (IHD), are one of the main cause of mortality and morbidity worldwide and are currently one of the top ten causes of death. Ischemic heart disease is a type of heart disease that is caused by narrowing of arteries feeding the heart itself. The present study aimed to use data mining algorithms in screening and early prediction of IHD according to the patient's characteristics and risk factors.

Methods: In this research, data of the first phase of Yazd Health Study (YaHS), focusing on 21 characteristics of 10,000 participants aged 20-70 years such as age, type of chest pain, blood sugar level, body mass index, employment status, etc. which have been collected since 2013 were analyzed.

Results: Data analysis was conducted using Random Forest and Naive Bayes algorithms which showed 74.51% accuracy in predicting IHD.

Conclusion: The study findings revealed that via applying Random Forest and Naive Bayes algorithms, ischemic heart disease can be predicted with high accuracy. Moreover, early screening and timely treatment in the early stages of disease may reduce mortality and morbidity.