A Low Cost Model for Diagnosing Coronary Artery Disease Based On Effective Features

Hamideh Ganji Arjenaki; Mohammad Hossein Nadimi Shahraki; Nasim Nourafza

Coronary artery disease which is one of the most dangerous diseases in the heart diseases field causes the death of many people every year. The diagnostic methods for this disease have high cost and many side effects on the patients. Even in the patients who have the similar signs of that disease, the use of these methods has high cost and unrelated side effects on them. Recently, researchers use data mining techniques to reduce the diagnosis cost of coronary artery disease. In this paper, firstly, the problem of cost of diagnosis of this disease is explained. Then, a model is proposed based on features selection to reduce the cost of diagnosis. The proposed model makes use of both genetic algorithm and perceptron neural network to select the most effective features. Then, the disease is diagnosed by Naïve Bayes classifier. Consistently, using these selected features reduces the cost. Although it is possible that the accuracy is decreased, the experimental results show that, the proposed model can maintain the accuracy.

Select Volume / Issues:

• Volume 6 Issue 1, Jan. 2015

Year:

2015

Type of Publication:

Article

Keywords:

Disease Diagnosis; Coronary Artery Disease; Medical Data Mining; Feature Selection

Journal:

IJECCE

Volume:

6

Number:

1

Pages:

93-97

Month:

Jan.-Feb.

ISSN:

2249-071X