Advertisement

Algorithms for Wave Form Classification

  • Chr. Zywietz
Chapter
  • 29 Downloads
Part of the Developments in Cardiovascular Medicine book series (DICM, volume 37)

Abstract

Algorithms for wave form classification serve two main purposes a) signal typing and b) diagnostic allocation of signal forms to diagnostic groups.

Wave form typing algorithms can be designed for interactive (supervised) learning or for self adaptive operation. Interactive learning procedures can be applied effectively in cases where only a few signal types occour and where a large number of forms has to be classified, i.e. in Holter ECGs or in monitoring. The performance of classification algorithms can often be improved by use of multidimensional feature vectors and transformed variables using their inner interdependence to enlarge the inter-group distance. For diagnostic allocation algorithms decision three type or multivariate classificators, i.e. the Bayes-formula are available. While morphology classification of signal forms can be done in a pure syntactical way, the development of diagnostic allocation algorithms needs, that the groups into which the signal shall be classified have to be defined by signal independent criteria from outside.

Keywords

Discriminant Function Wave Form Signal Form Discriminant Function Analysis Validation Logic 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Klusmeier, S. et al. 1978. Multivariate Alternativklassifikation, ein Versuch zur Überwindung einiger Nachteile der Bayes Klassifikation. Inf. Syst. i.d. Med. Versorg., 729–736, Schatthauser Verlag Stuttgart, New York 1978, Reichertz, Schwarz Edits.Google Scholar
  2. Lachenbruch, P.A. 1975. Discriminant Analysis, Hafner Press, A Division of Macmillan Publ. Co., Inc. New York, N.Y. 10022Google Scholar
  3. Swenne, C.A. et al. 1973. Adaption of Parameters for Interactive Coronary Care, Digest of the Xth Int. Conf. on Med. and biolog. Engin. Dresden 1973, p37Google Scholar
  4. Zywietz, Chr. et al. 1977. A New VCG Analysis Program for Children with multivariate Diagn. Classification. Computers in Cardiology 1977, 95–99, IEEE CATAL. Nr. 77CH1254-2CGoogle Scholar
  5. Zywietz, Chr. et al. 1981. HES LKG, a New Program for Computer Assisted Analysis of Holter Electrocardiograms. Computers in Cardiology 1981, 169–172, ISSN No. 0276-6574Google Scholar

Copyright information

© ECSC, EEC, EAEC, Brussels-Luxembourg 1984

Authors and Affiliations

  • Chr. Zywietz
    • 1
  1. 1.Arbeitsbereich Biosignalverarbeitung im Zentrum Bianetrie, Medizinische Informatik und MedizintechnikMedizinische Hochschule HannoverHannover 61Bundesrepublik Deutschland

Personalised recommendations