Novel CAD diagnosis method based on search, PCA, and AdaBoostM1 techniques

Eyupoglu, Can and Karakuş, Oktay ORCID: https://orcid.org/0000-0001-8009-9319 2024. Novel CAD diagnosis method based on search, PCA, and AdaBoostM1 techniques. Journal of Clinical Medicine 13 (10) , 2868. 10.3390/jcm13102868

PDF - Published Version Download (1MB)

Abstract

Background: Cardiovascular diseases (CVDs) are the primary cause of mortality worldwide, resulting in a growing number of annual fatalities. Coronary artery disease (CAD) is one of the basic types of CVDs, and early diagnosis of CAD is crucial for convenient treatment and decreasing mortality rates. In the literature, several studies use many features for CAD diagnosis. However, due to the large number of features used in these studies, the possibility of early diagnosis is reduced. Methods: For this reason, in this study, a new method that uses only five features—age, hypertension, typical chest pain, t-wave inversion, and region with regional wall motion abnormality—and is a combination of eight different search techniques, principal component analysis (PCA), and the AdaBoostM1 algorithm has been proposed for early and accurate CAD diagnosis. Results: The proposed method is devised and tested on a benchmark dataset called Z-Alizadeh Sani. The performance of the proposed method is tested with a variety of metrics and compared with basic machine-learning techniques and the existing studies in the literature. The experimental results have shown that the proposed method is efficient and achieves the best classification performance, with an accuracy of 91.8%, ever reported on the Z-Alizadeh Sani dataset with so few features. Conclusions: As a result, medical practitioners can utilize the proposed approach for diagnosing CAD early and accurately.

Item Type:	Article
Date Type:	Published Online
Status:	Published
Schools:	Computer Science & Informatics
Additional Information:	License information from Publisher: LICENSE 1: URL: https://creativecommons.org/licenses/by/4.0/, Type: open-access
Publisher:	MDPI
Date of First Compliant Deposit:	10 June 2024
Date of Acceptance:	7 May 2024
Last Modified:	10 Jun 2024 08:45
URI:	https://orca.cardiff.ac.uk/id/eprint/169632

Actions (repository staff only)

Edit Item