基於影像處理方法之心電圖小波係數研究

Abstract

本研究的目的在於以不同的量化方法來分析禪坐心電訊號(ECG)的小波係數(continuous wavelet coefficients, CWT)。研究中所採用的量化的方法包括：不變矩分析(invariant-moment analysis)、奇異值分解(singular value decomposition, SVD)、相關係數(correlation coefficient)以及變異數分析(ANOVA)。本研究的實驗受測者共有17位，實驗組是8位有禪坐經驗的受測者，而控制組是9位沒有禪坐經驗的受測者，兩組受測者的年齡相近。研究結果顯示，控制組的7個不變矩數值下降，而實驗組則是上升。在相關係數分析的部份，除了控制組的一位受測者之外，其餘受測者的心電訊號小波係數的主成份，在不同狀態下皆呈現很高的相關性。變異數分析的結果則顯示，比起實驗組，控制組的心電訊號小波係數在不同狀態下的差異較為顯著。因此，由以上結果可以得知實驗組的心電訊號波形較控制組穩定。

The aim of this research was to quantify the continuous wavelet coefficients (CWT) of raw ECG data. Several methods employed in this thesis included invariant-moment analysis, singular value decomposition (SVD), correlation coefficient, and analysis of variance (ANOVA). This study included 17 subjects, 8 experimental subjects with Zen-meditation experience and 9 control subjects in the same age range, yet, without any meditation experience. According to our results, the seven invariant moment values in control group tended to decrease, while the experimental group showed the tendency of increase. SVD analysis gives us another perspective. The correlation coefficients between major components of both groups showed a high value of correlation, although one result from the control group was considered to be moderately correlated. In ANOVA, differences appeared to be more significant in the control group than the experimental group. Thus, we may preliminarily suggest that ECG waveform patterns of experimental group behave more stably than those of control group in certain condition.