EURASIP Journal on Applied Signal Processing 
Volume 2006 (2006), Article ID 18268, 12 pages
doi:10.1155/ASP/2006/18268

Arrhythmic Pulses Detection Using Lempel-Ziv Complexity Analysis

Lisheng Xu,1 David Zhang,2 Kuanquan Wang,1 and Lu Wang1

 1Department of Computer Science and Engineering, School of Computer Sciences and Technology, Harbin Institute of Technology (HIT), Harbin 150001, China
2Department of Computing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China
Received 24 January 2005; Revised 9 September 2005; Accepted 12 September 2005

Recommended by William Sandham

Abstract

Computerized pulse analysis based on traditional Chinese medicine (TCM) is relatively new in the field of automatic physiological signal analysis and diagnosis. Considerable researches have been done on the automatic classification of pulse patterns according to their features of position and shape, but because arrhythmic pulses are difficult to identify, until now none has been done to automatically identify pulses by their rhythms. This paper proposes a novel approach to the detection of arrhythmic pulses using the Lempel-Ziv complexity analysis. Four parameters, one lemma, and two rules, which are the results of heuristic approach, are presented. This approach is applied on 140 clinic pulses for detecting seven pulse patterns, not only achieving a recognition accuracy of 97.1% as assessed by experts in TCM, but also correctly extracting the periodical unit of the intermittent pulse.