Lempel-ziv decomposition based arrhythmic pulses recognition

L Wang; K Wang; L S Xu

doi:10.1109/IEMBS.2005.1615496

Lempel-ziv decomposition based arrhythmic pulses recognition

Conf Proc IEEE Eng Med Biol Soc. 2005:2005:4606-9. doi: 10.1109/IEMBS.2005.1615496.

Authors

L Wang¹, K Wang, L S Xu

Affiliation

¹ Department of Computer Science and Technology, Harbin Institute of Technology (HIT), Harbin, China (phone: 86-451-89913312; e-mail: wanglu@hit.edu.cn).

PMID: 17281266
DOI: 10.1109/IEMBS.2005.1615496

Abstract

The computerization of traditional Chinese pulse diagnosis (TCPD) is relatively new in the field of automatic physiological signal analysis and diagnosis. The classification of pulse patterns according to their positions and shapes have been intensively investigated, but until now no research in identifying pulses by their rhythms has been found. This paper introduces a method to distinguish rhythmic and arrhythmic pulse patterns and further, applies the Lempel-Ziv decomposition to classify arrhythmic pulses. In the experiment on 140 clinic pulses, our approach was able to classify pulses by their rhythms with an accuracy of 90.7%.