Journal of Computer and Communications

Volume 6, Issue 10 (October 2018)

ISSN Print: 2327-5219   ISSN Online: 2327-5227

Google-based Impact Factor: 1.12  Citations  

An Efficient Arrhythmia Detection Using Autocorrelation and Statistical Approach

HTML  XML Download Download as PDF (Size: 2550KB)  PP. 63-81  
DOI: 10.4236/jcc.2018.610006    1,100 Downloads   2,794 Views  Citations

ABSTRACT

Computational electrocardiogram (ECG) analysis is one of the most crucial topics in cardiovascular research domain especially in identifying abnormalities of heart condition through cardiac arrhythmia symptom. There are many existing works focusing on recognizing the abnormalities condition through arrhythmia symptom, however, the detection rate is still unsatisfied. Arrhythmia consists of more than 14 various types of symptoms. Therefore, most of the existing research found it difficult to classify the entire symptom and maintain the overall accuracy especially in long hour data. In this study, a new mechanism to overcome this issue is proposed: A combination between Autocorrelation methods with K-Nearest Neighbor (KNN) classifier method is introduced to accurately and robustly detect 14 types of Arrhythmia symptom regardless of the origin of the symptom in a long hour data. Moreover, variability analysis based on periodic autocorrelation result is proposed and used for classification procedure. 1 minute and 12 hours duration data was chosen to compare and signify the most suitable time duration to detect Arrhythmia symptom. In addition, an analytical result and discussion is done to provide justification behind each tendency of Arrhythmia and Normal Sinus symptom in autocorrelation result. As the result of proposed method performance evaluation, it was revealed that the accuracy of 95.5% in discriminating Arrhythmia from Normal Sinus data is achieved. Furthermore, it was confirmed that utilizing autocorrelation result in long hour data can help to generalize abnormalities characteristic of heart condition like Arrhythmia symptom. It is concluded that the proposed method can be useful to diagnose abnormalities of heart condition at any stage.

Share and Cite:

Sinal, M. and Kamioka, E. (2018) An Efficient Arrhythmia Detection Using Autocorrelation and Statistical Approach. Journal of Computer and Communications, 6, 63-81. doi: 10.4236/jcc.2018.610006.

Cited by

No relevant information.

Copyright © 2024 by authors and Scientific Research Publishing Inc.

Creative Commons License

This work and the related PDF file are licensed under a Creative Commons Attribution 4.0 International License.