Open AccessFoetal ECG extraction by support vector regression
Author(s) -
Wei Zheng,
Xiaolong Li,
Xueyun Wei,
Hongxing Liu
Publication year - 2016
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
eISSN - 1350-911X
pISSN - 0013-5194
DOI - 10.1049/el.2016.0171
Subject(s) - support vector machine , regression , extraction (chemistry) , computer science , regression analysis , pattern recognition (psychology) , artificial intelligence , mathematics , statistics , chromatography , chemistry
A new method to extract the foetal electrocardiogram (ECG) using an abdominal ECG signal and a thoracic ECG signal of the mother is presented. The nonlinear mapping of the maternal ECG component from thoracic signal to an abdominal signal is approximated by the support vector regression (SVR) method. The foetal ECG signal is then extracted by subtracting the mapped thoracic signal from an abdominal signal. In an SVR algorithm, the nonlinear regression problem in an original signal space is transformed into a linear regression problem in high‐dimensional Hilbert space by kernel technology. It avoids the ‘overfitting’ problem and could obtain good results for the small sample training set. The validity of the proposed method is validated by the real ECG signals, and compared with the existing method.