Loading...

Noninvasive fetal ECG extraction using doubly constrained block-term decomposition

Mousavian, I ; Sharif University of Technology | 2020

509 Viewed
  1. Type of Document: Article
  2. DOI: 10.3934/mbe.2020008
  3. Publisher: American Institute of Mathematical Sciences , 2020
  4. Abstract:
  5. Fetal electrocardiogram (fECG) monitoring is a beneficial method for assessing fetal health and diagnosing the fetal cardiac condition during pregnancy. In this study, an algorithm is proposed to extract fECG from maternal abdominal signals based on doubly constrained block-term (DoCoBT) tensor decomposition. This tensor decomposition method is constrained by quasi-periodicity constraints of fetal and maternal ECG signals. Tensor decompositions are more powerful tools than matrix decomposition, due to employing more information for source separation. Tensorizing abdominal signals and using periodicity constraints of fetal and maternal ECG, appropriately separates subspaces of the mother, the fetus(es) and noise. The quantitative and qualitative results of the proposed method show improved performance of DoCoBT decomposition versus other tensor and matrix decomposition methods in noisy conditions. © 2020 the Author(s)
  6. Keywords:
  7. BT decomposition ; fECG ; Electrocardiography ; Fetal monitoring ; Source separation ; Block term decompositions ; Fetal ECG extractions ; Fetal electrocardiograms ; Matrix decomposition ; Quasi-periodicities ; Signal extraction ; Tensor decomposition ; Tensors ; Algorithm ; Controlled study ; Decomposition ; Female ; Fetus ; Fetus electrocardiography ; Human ; Noise ; Periodicity ; Quantitative analysis ; ECG
  8. Source: Mathematical Biosciences and Engineering ; Volume 17, Issue 1 , 2020 , Pages 144-159
  9. URL: https://www.aimspress.com/article/10.3934/mbe.2020008