A Reliable Arrhythmias-Recognition Scheme Via Wavelet and Multiclass Support Vector Machine
Magdi B. M. Amien1, Tamadur El-Khansa Japir2
1Magdi B. M. Amien, Dept. of Electronics Engineering, Faculty of Engineering & Technology, University of Gezira, Wed-Madani, Sudan.
2Tamadur El-Khansa Japir, Dept. of Biomedical Engineering, Faculty of Engineering, Sudan University of Science & Tech. Khartoum, Sudan.
Manuscript received on June 25, 2014. | Revised Manuscript received on July 03, 2014. | Manuscript published on July 05, 2014. | PP: 119-123  | Volume-4, Issue-3, July 2014. | Retrieval Number: C2314074314 /2012©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Heart diseases (HD) are the number one cause of death globally, more people die annually from HDs than from any other cause, according to World-Health-Organization (WHO) 7.3 million were died due to coronary heart disease in 2008. Electrocardiogram (ECG) interpretation is most widely used to detect the abnormality of the heart. A reliable computer programs could lead to enhanced visual interpretation, and significant-increase of diagnosis-efficiency. This study introduced a novel method for ECG classification; fifteen different records of five rhythms from “MIT-BIH” Arrhythmia Database have been used to evaluate the implemented algorithms. The proposed approach consists of three distinct stages. In the first stage a preprocessing of different-steps is done to remove the baseline wander, power line interference and to enhance morphological properties. Secondly Daubechies are chosen and implemented as mother-wavelet-function to extract ten features of ECG signals, in the final stage Support-Vector-Machines (SVM), has been used as Multi-class classifier and decision making algorithm. The performance of the proposed method has bees evaluated in terms of accuracy, and specific accuracy. The experimental results have shown that the proposed system achieves validity as competitive results quality-wise, and the accuracy-rate of classification of Normal sinus Rhythm (N), Bundle Branch Block (RBBB), Atrial Premature Beat (APB), 3Premature Ventricular Contraction (PVC), Fusion Heart Beats (F), and Unclassified Heart Beats (P) were 90.0%, 100%, 66.6%, 100%, 100%, and 100%, respectively.
Keywords: Arrhythmias classification, Wavelet, MSVM