• DocumentCode
    169986
  • Title

    Segmentation of nocturnal Holter-ECG recordings with respect to different recumbent body positions

  • Author

    Maier, Christian ; Benz, Alexander ; Dickhaus, H.

  • Author_Institution
    Univ. of Heidelberg, Heidelberg, Germany
  • fYear
    2014
  • fDate
    25-28 May 2014
  • Firstpage
    159
  • Lastpage
    160
  • Abstract
    This paper evaluates a previously developed approach for minute-by-minute segmentation of night-time ECG recordings with respect to periods of different recumbent body positions (BPs). It is based on recurrence analysis of 8D-feature vectors aggregating the area under the QRS complex in eight Holter-ECG leads. The number and consistency of ECG-derived BPs was compared to reference BPs (45° resolution) extracted from 139 polysomnograms (PSGs, 120 subjects). The results indicate a high BP resolution for the ECG. Up to 3 ECG BPs per PSG BP were identified in 70% of the cases. In 60% of the ECGs, the fraction of inconsistent epoch assignments was ≤10%, and ≤30% of BP changes in the PSG were undetected. When adjacent PSG BPs were considered exchangeable, the results improved significantly, indicating a potentially unfavorable effect of the PSG sensor discretization and encouraging further steps towards the absolute classification of BPs at least in coarse categories.
  • Keywords
    electrocardiography; medical signal processing; signal classification; 8D-feature vectors; ECG BP; PSG BP; PSG sensor discretization; QRS complex; epoch assignments; night-time ECG recordings; nocturnal holter-ECG recording segmentation; polysomnograms; recumbent body positions; recurrence analysis; Electrocardiography; Europe; Feature extraction; Monitoring; Oscillators; Sleep apnea; Vectors;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Cardiovascular Oscillations (ESGCO), 2014 8th Conference of the European Study Group on
  • Conference_Location
    Trento
  • Type

    conf

  • DOI
    10.1109/ESGCO.2014.6847568
  • Filename
    6847568