A DECISION TREE BASED APPROACH FOR THE IDENTIFICATION OF ISCHAEMIC BEATS IN ECG RECORDINGS

• Published in: Mathematical Methods In Scattering Theory And Biomedical Engineering pp. 312 - 319
• Main Authors: EXARCHOS, T. P., PAPALOUKAS, C., FOTIADIS, D. I.
• Format: Book Chapter
• Language: English
• Published: WORLD SCIENTIFIC 01.08.2006
• Subjects: Biomedical Engineering
• ISBN: 9812568603; 9812773193; 9789812773197; 9789814477598; 9814477591; 9789812568601
• DOI: 10.1142/9789812773197_0031

Cardiac beat classification is a key process in the detection of myocardial ischaemic episodes in the electrocardiographic (ECG) signal. In this, study we propose an automated methodology for the classification of cardiac beats in long duration ECG recordings. The proposed approach is based on the decision tree induction algorithm and consists of three stages. In the first stage the ECG signal is preprocessed and noise removal takes place. Baseline wandering, electromyographic contamination and A/C interference are eliminated. In the second stage five representative features are extracted from every cardiac beat. These features are measurements from the ST segment and the T wave. Another demographic feature, the patient's age is also employed. In the last stage of the method, the decision tree induction algorithm is used. A decision tree is induced from the training set and it is employed in order to classify the test set in two categories: normal and ischaemic cardiac beats. For the evaluation of our methodology, a cardiac beat dataset was constructed using several recordings of the European Society of Cardiology ST-T database. The obtained sensitivity (Se) and specificity (Sp) was 87% and 92%, respectively.