DocumentCode :
3106764
Title :
Real-Time ECG Signal Feature Extraction for the Proposition of Abnormal Beat Detection - Periodical Signal Extraction
Author :
Szczepanski, Adam ; Saeed, Khalid
Author_Institution :
Fac. of Phys. & Appl. Comput. Sci., AGH Univ. of Sci. & Technol., Kraków, Poland
fYear :
2013
fDate :
5-7 July 2013
Firstpage :
261
Lastpage :
266
Abstract :
This paper presents a novel algorithm for ECG signal feature extraction and proposes a solution for detection of abnormal beats. The main goal is to extract signal and heartbeat parameters by using local extreme values and their dependences. This produces the averaged heartbeat which is used to analyze the signal in real time to detect anomalies. The main advantages of our approach are fast computation and portability between software development platforms. The main disadvantage is the need of clear fragments of samples for the training of the algorithm. Experimental results indicate the proposed system is efficient when compared to other models.
Keywords :
electrocardiography; feature extraction; medical signal detection; abnormal beat detection; anomaly detection; heartbeat parameters; local extreme values; periodical signal extraction; real-time ECG signal feature extraction; signal analysis; software development platform portability; Algorithm design and analysis; Electrocardiography; Heart beat; Medical services; Real-time systems; Software; Software algorithms; Biomedical signal processing; Electrocardiography; Signal analysis; Signal processing algorithms;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Biometrics and Kansei Engineering (ICBAKE), 2013 International Conference on
Conference_Location :
Tokyo
Type :
conf
DOI :
10.1109/ICBAKE.2013.51
Filename :
6603513
Link To Document :
بازگشت