Comparative Study of ECG Classification Performance Using Decision Tree Algorithms

Comparative Study of ECG Classification Performance Using Decision Tree Algorithms

Faiza Charfi (Electronic and Information Technologies Laboratory, Sfax University, Sfax, Tunisia) and Ali Kraiem (Electronic and Information Technologies Laboratory, Sfax University, Sfax, Tunisia)
Copyright: © 2012 |Pages: 19
DOI: 10.4018/jehmc.2012100106

Abstract

The electrocardiogram (ECG) signal has often been reported to play an important role in the primary diagnosis, prognosis, and survival analysis of heart diseases. Electrocardiography has brought several valuable impacts on the practice of medicine. This paper deals with the feature extraction and automatic analysis of different ECG signal waves using derivative based/ Pan-Tompkins based algorithms. The ECG signal contains an important amount of information that can be exploited in different way. It allows for the analysis of cardiac health condition. The discrimination of ECG signals using the Data Mining Decision Tree techniques is of crucial importance in the cardiac disease therapy and control of cardiac arrhythmias. Different ECG signals from MIT/BIH Arrhythmia data base are used for ECG features extraction and analysis. Two pathologies are considered: atrial fibrillation and right bundle branch block. Some decision tree classification algorithms currently in use, including C4.5, Improved C4.5, CHAID (Chi square Automatic Interaction Detector) and Improved CHAID are performed for performance analysis. Promising results have been achieved using the C4.5 classifier, with an overall accuracy of 96.87%.
Article Preview

1. Introduction

The ECG signal analysis is extensively used in the cardiac pathology diagnoses. The research on pathology consists in detecting and identifying the different waves constituting the ECG signal, to measure their lengths as well as their amplitudes and in short to establish a diagnosis.

The ECG is a recording of the heart’s electrical activity. It provides valuable information about heart functional aspects. The ECG waveform is divided into P, Q, R, S, T, and U elements (Saad, Abdullah, & Low, 2006). P wave represents the atrial depolarization that shows the contraction of the left and right atria. Its duration is between 0.06 to 0.12 seconds for a normal contraction. The QRS complex represents ventricular activation or depolarization (Guler & Ubeyli, 2005). The QRS complex duration is less than 0.1 seconds for normal ventricles contraction. The T wave represents ventricles’ depolarization which set up the cardiac muscle for another contraction. The PR interval is the conduction time required for an electrical impulse to be conducted from the atria to the ventricles. The duration is normally 0.12 to 0.20 seconds and is used to diagnose heart block problems (Saad, Abdullah, & Low, 2006). The ST segment is an isoelectric. It represents the period when the ventricles are depolarized. It is measured from the end of the S wave until the beginning of the T wave. Whereas, the ST interval is defined from the S wave end to the T wave end Figure 1. The normal value of heart beat ranges from 60 to 100 beats/mn. The ECG strips are best interpreted from lead II or lead VI which shows the most clearly rhythm of the heart according to Einthoven’s Triangle (Saad, Abdullah, & Low, 2006).

Figure 1.

ECG signal

Deviation and distortion in any parts of ECG that is called Arrhythmia can illustrate a specific heart disease. Atrial fibrillation (AF) represents one of the most current cardiac arrhythmias and corresponds to the dysfunction of atrial. It is the result of disorganization in the atrial electric activity. The P wave analysis is therefore very important in the case of subjects with atrial fibrillation risk. Whereas right bundle branch block (RBBB) corresponds to the deterioration of atrio-ventricular conduction in the heart right side during the chronic phase of the disease. The duration criterion for the QRS-complex and PR interval in RBBB are very important.

Hence, the ECG interpretation is important for cardiologists to decide the diagnostic categories of cardiac problems. In the case of problems which are a matter of pattern recognition, the need is to use reliable methods that maintain the data structure, that do not call for very high statistical hypotheses, and that provide models easy to interpret (Chazal & Reilly, 2006). Among the techniques which correspond best to these characteristics, the Data Mining takes an important place. Data Mining is an iterative process within which progress is defined by discovery, either through automatic or manual methods. Data Mining is most useful in an exploratory analysis scenario in which there are no predetermined notions about what will constitute an interesting outcome. The Data Mining classification techniques were used efficiently in detecting fraudulent financial statements and identifying the associated factors (Kirkos, Charalambos, & Manolopoulos, 2007). Their application for physiological signal classification in combination with the segmentation methods is a new approach for detecting any abnormal activities.

In this study, we carry out an in-depth examination of publicly available ECG Arrhythmias database (Moody & Mark, 1990) in order to detect some ECG abnormalities by using Data Mining classification methods. This work is motivated by developing an accurate and fast method for real-time ECG analysis which is mandatory in fully automated monitoring devices for cardiac patients. We used the decision tree algorithms of Data Mining techniques in differentiating between some clinical and pathological observations and identifying variables that mostly affect the ECG.

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 10: 4 Issues (2019): Forthcoming, Available for Pre-Order
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing