Myocardial Ischemia Detection by Pulse Signal Feat

Myocardial Ischemia Detection by Pulse Signal Features and Fuzzy Clustering

Kang-Ming Chang1, Zhi-Zhong Lin1 ,Shing-Hong Liu2 ,Chu-Chang Tyan3
1Department of Computer and Communication Engineering, Asia University, Taiwan
2Department of Biomedical Engineering, Yuanpei University, Taiwan
3Department of Chinese Medicine, China Medical University Hospital, Taiwan
changkm@asia.edu.tw, keigo.o@yahoo.com.tw, shliu@mail.ypu.edu.tw,
tyan@www.cmuh.org.tw

Abstract
The aim of this study is to propose a new diagnosing algorithm, for myocardial ischemia, based
on the newly developed non-invasive measurement, three-axis sphygmography, and further processed by
modified fuzzy c-means algorithms. The three-axis sphygmography measures pressure pulse waveform
(PPW), and the PPW are analyzed by its harmonic components in the frequency domain. The first three
harmonics of the Fourier series of PPW are significantly different between myocardial ischemia
patients and normal subjects. The harmonic of the Fourier series of PPW are feed into a modified fuzzy cmeans
algorithm, and developed PPW criteria to identify myocardial ischemia with sensitivity 69% and
specificity 100%.