A Ventricular Far-Field Artefact Filtering Technique for Atrial Electrograms

Simanto Saha, Simon Hartmann, Dominik Linz, Prashanthan Sanders, Mathias Baumert

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Intracardiac atrial electrograms (EGM) are prone to ventricular far-field potentials due to ventricular depolarization. In this study, a filtering technique integrating independent component analysis (ICA) and wavelet decomposition has been proposed to significantly reduce the ventricular far-field contents while preserving the EGM morphology related to atrial activations. First, the wavelet decomposition is applied to each unipolar EGM. Then, ICA is applied to the decomposed unipolar EGM components and surface ECG template. Each independent component is cross-correlated with the simultaneously recorded ECG template and the three components with higher correlation coefficients were eliminated before applying inverse ICA. Total of 126 unipolar EGM collected from an atrial fibrillation patient have been included. Results indicate that the proposed filtering can reduce the ventricular signal power by around 17 dB (decibel). Furthermore, the signal-to-noise ratio is increased by approximately 17 dB after applying the proposed filtering. In conclusion, the proposed filtering method could be used for atrial fibrillation-related intracardiac mapping for catheter ablation. Further studies on a larger dataset are essential to quantify the exact impact of ventricular artefacts on both unipolar and bipolar EGM and the effectiveness of the proposed filtering technique.

Original languageEnglish
Title of host publication2019 Computing in Cardiology, CinC 2019
PublisherIEEE Computer Society
ISBN (Electronic)9781728169361
DOIs
Publication statusPublished or Issued - Sep 2019
Externally publishedYes
Event2019 Computing in Cardiology, CinC 2019 - Singapore, Singapore
Duration: 8 Sep 201911 Sep 2019

Publication series

NameComputing in Cardiology
Volume2019-September
ISSN (Print)2325-8861
ISSN (Electronic)2325-887X

Other

Other2019 Computing in Cardiology, CinC 2019
CountrySingapore
CitySingapore
Period8/09/1911/09/19

Keywords

  • Cardiology
  • Electrocardiography
  • Morphology
  • Atrial electrograms
  • Atrial fibrillation
  • Intracardiac mapping
  • Biomedical signal processing

ASJC Scopus subject areas

  • Computer Science(all)
  • Cardiology and Cardiovascular Medicine

Cite this