DocumentCode
1778525
Title
Fetal electrocardiogram extraction from maternal abdominal signals
Author
Ivanushkina, N. ; Ivanko, K. ; Lysenko, E. ; Borovskiy, I. ; Panasiuk, O.
Author_Institution
Kyiv Polytech. Inst., Nat. Tech. Univ. of Ukraine, Kiev, Ukraine
fYear
2014
fDate
15-18 April 2014
Firstpage
334
Lastpage
338
Abstract
An integrated approach for fetal electrocardiogram (ECG) extraction based on the joint application of wavelet transform (WT) and independent component analysis (ICA) is proposed. This research is aimed at development of computer based electrocardiography systems of high resolution, which require the advanced methods for early noninvasive diagnostics of fetal status. The results of this study demonstrate that ICA can effectively distinguish the fetal independent components, but requires the use of additional methods for noise removing from these components. Elimination of noise is achieved by WT applying. Joint application of ICA and WT allowed us to extract with slight distortions of useful signal the simulated fetal ECG from the artificial signal containing maternal ECG and noise.
Keywords
electrocardiography; feature extraction; independent component analysis; medical signal processing; obstetrics; signal denoising; signal resolution; wavelet transforms; artificial signal; computer based electrocardiography systems; fetal electrocardiogram extraction; fetal independent components; independent component analysis; integrated approach; joint application; maternal ECG; maternal abdominal signals; noise elimination; noise removing; noninvasive diagnostics; simulated fetal ECG; wavelet transform; Electrocardiography; Noise; Shape; Source separation; Wavelet analysis; Wavelet transforms; abdominal electrocardiogram; fetal electrocardiogram; independent component analysis; modeling; monitoring of fetal condition; wavelet transform;
fLanguage
English
Publisher
ieee
Conference_Titel
Electronics and Nanotechnology (ELNANO), 2014 IEEE 34th International Conference on
Conference_Location
Kyiv
Print_ISBN
978-1-4799-4581-8
Type
conf
DOI
10.1109/ELNANO.2014.6873910
Filename
6873910
Link To Document