The prediction model for cardiovascular disease using Yazd's health study data (YaHS)

Abstract

Introdution: Ischemic heart disease is one of the most common diseases, which has led to high mortality rates all over the world. This disease is caused by narrowing or blockage of coronary arteries, which are the provider of blood to the heart. Identifying the people susceptible to this disease and bringing changes in their lifestyles has been said to reduce the related mortality rates and increase the patient's longevity.

Methods: Yazd people Health Study (YaHS) was conducted on a random sample of 10,000 people living in the city of Yazd, Iran in the years 2014-15 for a general health and disease survey. These data were first balanced by bootstrapping technique due to their unbalanced nature. Next, classification methods were used in the training phase. Various classifiers, such as artificial neural network, rule inducer, regression, and AdaBoost were used in order to evaluate the proposed method with two scenarios.

Results: The results showed that the screening of the people susceptible to ischemic heart disease had the most significant effect on increasing the sensitivity of the discovery classifier of CN2 subgroup through using balanced data by bootstrapping method followed by their analysis for the purpose of producing a sample of the patients. This classifier proved to have the potential for detecting 83.6% of the people susceptible to this disease.

Conclusion: Therefore, it can be concluded that data mining methods are effective in screening for susceptible people with ischemic heart disease. This method can be compared with other traditional screening methods in that it is more cost-effective and faster.