Conference Paper

Detection of events in biomedical signals by a Rényi entropy measure

Conference Proceeding cp
AIP Conference Proceedings
  • Volumen: 860
  • Fecha: 01 December 2006
  • Páginas: 210-219
  • ISSN: 0094243X 15517616
  • ISBN: 0735403562
  • Source Type: Conference Proceeding
  • Document Type: Conference Paper
Biomedical signals contain important information about the healthy condition of human beings. Anomalous events in these signals are commonly associated to diseases. The information content enclosed by time-frequency representations (TFR) of biomedical signals can be explored by means of different Rényi entropy measures. To be precise, Rényi entropy can be approached under different normalizations, producing different outcomes. The best choice depends upon the particularities of the application considered. In this paper we propose a new processing scheme to the problem of events detection in biomedical signals, based on a particular normalization of the Rény entropy measurement. As in the case of another TFR's, the pseudo-Wigner distribution (PWD) of a biomedical signal can take negative values and thus it cannot be properly interpreted as & probability density function. Therefore a complexity measure based on the classical Shannon entropy cannot be used and a generalized measure such as the Rényi entropy is required. Our method allows the identification of the events as the moments having the highest amount of information (entropy) along the temporal data. This provides localized information about normal and pathological events in biomedical signals. Therefore, the diagnosis of diseases is facilitated in this way. The method is illustrated with examples of application to phonocardiograms and electrocardiograms and result are discussed. © 2006 American Institute of Physics.

Author keywords

    Indexed keywords

      Funding details