摘要
本文通过引入分形概念,首次为描述心音信号的复杂程度给出了一个定量的指标——分数维数。研究结果表明:心音信号确实存在着“无标度区”,具有典型的分形特性;心音信号的分维值能显著性地区分正常心音、二尖瓣病变心音(包括二尖瓣狭窄、二尖瓣关闭不全及二尖瓣狭窄加关闭不全)和主动脉瓣病变心音(主动脉瓣狭窄、主动脉瓣关闭不全及主动脉瓣狭窄加关闭不全)。
<ABSTRACT>
Heart sound contains plenty of information in reference to the sta-
tes of heart and cardiovascular conditions. While the spectrogram of
heart sound have been playing an important role in clinical diagnosis,
sometimes it is difficult even for trained users to find out some differ-
ences on the spectrogram, which are perceptible to the ear. And there
is another great default on spectrogram analysis that some very differ-
ent shapes in time domain may become the same in frequency domain.
To find out new characterized parameters and capture the information
missing in the spectrogram, we introduced fractal concept into heart
sound analysis, and advnoed a new paramter, fractal dimension (D),
to characterize quantitatively the complexity of heart sound. Fractal
Oeometry, founded by B.B.Mandelbrot in 1973, provide a new qualita-
tive and quantitative approach for understanding the complex shapes of
nature. Heart sound is highly irregularly shaped signal, which motiva-
tes us to test its fractal characterization. The fractal analysis results
showed that there is indeed a range of 'scale-invariance'to heart sound,
and the statistical analysis results showed that the fractal dimensions
of heart sound can significantly distinguish the different states among
the normal (D = 1. 394±0. 044 ) and certain cardiac pathologies (mitral st-
enosis (D = 1. 768±0.005 ), mitral insufficiency (D = 1. 806±0.012 ), mit-
ral stenosis and insufficiency (D = 1. 607±0.026 ), aortic stenosis (D =
1 .657±0.023 ), aortic incompetence (D = 1. 859±0. 016 ), aortic stenosis
and incompetence (D = 1. 527±0.019 ).
出处
《中国生物医学工程学报》
EI
CAS
CSCD
北大核心
1993年第1期50-55,共6页
Chinese Journal of Biomedical Engineering
关键词
心音信号
分形
Heart sound
Fractals
