Annals of Biomedical Engineering

, Volume 22, Issue 5, pp 480–492 | Cite as

Localized spatial discrimination of epicardial conduction paths after linear transformation of variant information

  • Edward J. Ciaccio
  • Stanley M. Dunn
  • Metin Akay
  • Andrew L. Wit
  • James Coromilas
  • Constantinos A. Costeas
Research Articles

Abstract

We present a method for the localized statistical discrimination of class populations based on the Karhunen-Loève and Fukunaga-Koontz transforms. These transforms provide features that model the variance of a sample distribution. The spatial series of a 196 channel epicardial electrogram recording from an arrhythmogenic postinfarct canine were analyzed. For each type of rhythm studied, Karhunen-Loève and Fukunaga-Koontz expansions were computed from five training sets of spatial data, corresponding to five locations across the surface of the heart. Nonparametric statistical tests were then used for discriminant analysis to compare properties representative of the distribution from each proposed class. In a comparison of properties from sinus rhythm to those of two ventricular tachycardias, several spatial regions exhibited statistically significantly different propagation characteristics. These areas were observed by visual inspection of electrogram activation maps to be characterized by conductive gradients, which differed in magnitude and direction from one rhythm to another. The regions in which the propagation characteristics are of greatest difference in each tachycardia were centered upon sites of conduction block, manifested by reentrant circuit rhythms. Therefore, the importance of the technique for the localization of specific electrophysiologic events is demonstrated. This study extends previous work of our group on biosignal pattern recognition to encompass localized spatial data.

Keywords

Circuit rhythm Electrogram Fukunaga-Koontz transformation Karhunen-Loève transformation Nonparametric Pattern recognition Rank correlation Reentry 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Akay, M. Biomedical Signal Processing, New York: Academic Press, 1993, 377 pp.Google Scholar
  2. 2.
    Anderson, K. P., R. Walker, P. R. Ershler, M. Fuller, T. Dustman, R. Menlove, S. V. Karwandee, and R. L. Lux. Determination of local myocardial electrical activation for activation sequence mapping.Circ. Res. 69:898–917, 1993.Google Scholar
  3. 3.
    Bigun, J. Unsupervised feature reduction in image segmentation by local Karhunen-Loève transform. In: Proc. IEEE 11th IAPR Int. Conf. Pat. Rec., Vol. II, Edited by E. Backer and E. S. Gelsema. Los Alamitos, California; IEEE Comp Soc Press, The Hauge, The Netherlands, 1992, pp. 79–83.Google Scholar
  4. 4.
    Ciaccio, E. J. Representation and comparison of sample distributions for biomedical signal pattern recognition. Ph.D. thesis, Rutgers University, New Brunswick, New Jersey, 1993.Google Scholar
  5. 5.
    Ciaccio, E. J., S. M. Dunn, and M. Akay. Pattern recognition and interpretation systems for biomedical signal processing: fundamentals.IEEE-EMB Mag. 12:89–97, Sept., 1993.Google Scholar
  6. 6.
    Ciaccio, E. J., S. M. Dunn, M. Akay, and A. L. Wit. Discrimination of normal sinus rhythm versus ventricular tachycardia in the arrythmogenic canine heart after myocardial infarction. In: ES-NICVD 5th World Cong. Noninvasive Cardiovasc. Dyn. (abstract), New Brunswick, New Jersey, 1993.Google Scholar
  7. 7.
    Ciaccio, E. J., S. M. Dunn, M. Akay, A. L. Wit, and C. A. Costeas. The use of correlation analysis for determining spatiotemporal relationships of multichannel ventricular electrogram data in the infarcted canine. In: IEEE Proc. 13th Southern Biomedical Engineering Conf., pp. 441–444, April 16–17, 1994.Google Scholar
  8. 8.
    Ciaccio, E. J., S. Weiner, S. S. Reisman, S. M. Dunn, and M. Akay. Pattern recognition and interpretation of electromyogram data from cat jaw muscle.Comp. Biol. Med. 24:19–30, 1994.CrossRefGoogle Scholar
  9. 9.
    Cliff, A. D., and J. K. Ord. Spatial processes: Models and Applications. London: Pion Ltd., 1981, 266 pp.Google Scholar
  10. 10.
    Dillon, S. M. Techniques for studying reentry. In: Cardiac Electrophysiology: A Textbook, edited by M. R. Rosen M. J. Janse, and A. L. Wit. Mount Kisco, New York: Futura, 1990, pp. 537–553.Google Scholar
  11. 11.
    Dillon, S. M., M. A. Allessie, P. C. Ursell, and A. L. Wit. Influences of anisotropic tissue structure on reentrant circuits in the epicardial border zone of subacute canine infarcts.Circ. Res. 63:182–206, 1988.PubMedGoogle Scholar
  12. 12.
    Dunn, S. M. An introduction to model-based imaging.Dentomaxillofac. Radiol. 21:184–189, 1992.PubMedGoogle Scholar
  13. 13.
    Estola, K.-P., and M. Jokipii. Robust features for ECG pattern recognition algorithms. In: Proc. 1987. Comp. in Cardiol. Conf., Vol. 14, Edited by J. R. Cox, Jr. and J. Meyer. Washington, D.C.:IEEE Comp Soc Press, 1988, pp. 253–256.Google Scholar
  14. 14.
    Fukunaga, K., and W. L. G. Koontz. Application of the Karhunen-Loève expansion to feature selection and ordering.IEEE Trans. Comp. C-19:311–318, 1970.Google Scholar
  15. 15.
    Hays, W. L., and R. L. Winkler. Statistics: Probability, Inference, and Decision, Vol. II. New York, HRW, 1970, 315 pp.Google Scholar
  16. 16.
    Kanto, A. On the Estimation of the Inverse Autocorrelation Function. Ph.D. thesis, University of Tampere, Tampere, Finland, 1983.Google Scholar
  17. 17.
    Kozmann, G., L. S. Green, and R. L. Lux. Nonparametric identification of discriminitive information in body surface maps.IEEE Trans. Biomed. Eng. 38:1061–1068, 1991.CrossRefPubMedGoogle Scholar
  18. 18.
    Lux, R. L., Karhunen-Loève representation of ECG data.J. Electrocardiol. 25(Suppl.):195–198, 1992.PubMedGoogle Scholar
  19. 19.
    Lux, R. L., A. K. Evans, M. J. Burgess, R. F. Wyatt, and J. A. Abildskov. Redundancy reduction for improved display and analysis of body surface potential maps. I. Spatial compression.Circ. Res. 49:186–196, 1981.PubMedGoogle Scholar
  20. 20.
    Millar, C. K., F. A. Kralios, and R. L. Lux. Correlation between refractory periods and activation-recovery intervals from electrograms: effects of rate and adrenergic interventions.Circulation 72:1372–1379, 1985.PubMedGoogle Scholar
  21. 21.
    Press, W. H., S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery. Numerical Recipes in Fortran. New York: Cambridge University Press, 1992, 963 pp.Google Scholar
  22. 22.
    Spach, M. S., and J. M. Kootsey. Relating the sodium current and conduction to the shape of transmembrane and extracellular potentials by simulation: effects of propagation boundaries.IEEE Trans. Biomed. Eng. 32:743–755, 1985.PubMedGoogle Scholar
  23. 23.
    Watabe, S., H. Hayashi, Y. Yamada, K. Miyaji, S. Yabe, I. Sotobata, A. Iwata, and N. Suzumura. Application of the Karhunen-Loève expansion to evaluate regional cardiac excitation in body surface potential maps.J. Electrocardiol. 23:33–40, 1990.CrossRefPubMedGoogle Scholar
  24. 24.
    Wit, A. L.. Cellular electrophysiological mechanisms of cardiac arrythmias. In: Control of Cardiac Arrythmias by Lengthening Repolarization, edited by B. N. Singh. Mount Kisco, New York:Futura, 1988, pp. 1–31.Google Scholar
  25. 25.
    Wit, A. L., and M. J. Janse. The Ventricular Arrythmias of Ischemia and Infarction. Mount Kisco, New York, Futura, 1993, 648 pp.Google Scholar
  26. 26.
    Wit, A. L., and M. E. Josephson. Fractionated electrograms and continuous electrical activity: fact or artifact. In: Cardiac Electrophysiology and Arrythmias, edited by D. Zipes and J. Halife Orlando, Florida: Grune and Stratton, 1985, pp. 343–351.Google Scholar

Copyright information

© Biomedical Engineering Society 1994

Authors and Affiliations

  • Edward J. Ciaccio
    • 1
    • 2
  • Stanley M. Dunn
    • 1
    • 2
  • Metin Akay
    • 1
  • Andrew L. Wit
    • 3
  • James Coromilas
    • 3
  • Constantinos A. Costeas
    • 3
  1. 1.Department of Biomedical EngineeringRutgers UniversityPiscataway
  2. 2.Department of Oral and Maxillofacial RadiologyUMDNJ-NJ Dental SchoolNewark
  3. 3.Department of PharmacologyColumbia UniversityNew York

Personalised recommendations