Time-frequency approach to ECG classification of myocardial infarction. (June 2020)

• Record Type: Journal Article
• Authors: Kayikcioglu, İlknur; Akdeniz, Fulya; Köse, Cemal; Kayikcioglu, Temel
• Is Part Of: Computers & electrical engineering. Volume 84(2020)
• Publication Date: 2020-06
• Subjects: Myocardial infarction -- Electrocardiogram -- ST segment -- Classification -- Time-frequency distributions -- Telemedicine; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2020.106621
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: Time-frequency distributions based features are used to predetermine ST segment changes. A large database consisting of 335, 064 ECG R-R intervals are used in leads V1, V2, V3, V4 and V5. The results of classification with four classes (healthy, arrhythmia, ST segment depression and ST segment elevation) are found by many classification methods. The speed of the proposed algorithm is suitable for telemedicine systems. The accuracy of detecting the elevations or depression ST segment is found to be 94.23%. Abstract: Electrocardiogram (ECG) analysis is one of the most important techniques to classify myocardial infarction. It is possible to diagnose that the patient may have a heart attack with ST segment elevation or depression in the ECG recordings taken before patient has a myocardial infarction. We propose a method to classify ST segment using time-frequency distribution based features from multi-lead ECG signals. In contrast to many studies in the literature, the proposed method is based on four-class classifcation method and is tested on a large dataset consisting of three different databases, namely MIT-BIH Arrhythmia database, European ST-T database and Long-Term ST database. Among the classification algorithms, the weighted k-NN algorithm achieved the best average performance with accuracy of 94.23%, sensitivity of 95.72% and specificity of 98.15% using Choi-Williams time-frequency distribution features. Meanwhile, the speed of the proposed algorithm is suitable for telemedicine systems.