臺灣博碩士論文加值系統

在本篇論文中,將對心電圖中QRS波頂點R至T波頂點T之間距,簡稱R-T區間,
進行研究及分析。研究結果發現RT區間在經過一次差分之後,其ACF圖與一
階移動平均(MA(1))模型的ACF圖有相同的結構,所以可建立RT區間的一階
整合移動平均模型(ARIMA(0,1,1)),對於每筆資料中 ARIMA(0,1,1)擬合所
得到的轍數為一觀測值,然後對這些轍數利用變異數分析的方法來分析,發
現正常嬰兒的轍數比有病嬰兒的轍數顯著的低。 另外再參考陳宏
天(1996)所求的PR區間之轍數,其所求PR區間的轍數也是發現正常嬰兒比
有病嬰兒的轍數顯著的低,然後對此二者之轍數利用多變量分析的方法來
分析,發現對RT區間和PR區間來說,正常和有病的轍數是有顯著差異的。

In this thesis, we study the R-T interval of ECG, which is
defined as the interval between the adjacent peaks of QRS-wave
and T-wave. The result reveals that the autocorrelation
functions of the R-T interval after differencing have the same
structure as that of a moving average model of order one.
Therefore, we apply the method of the time domain in time
series to build an ARIMA(0,1,1) model, furthermore, the
estimated moving average coefficient is named "rut" for
simplicity. We make further research by using the method of
ANOVA to distinguish normal and abnormal children. Generally
speaking, the ruts for normal children are lower than sick
children. Besides,we also use MANOVA to analyze the ruts for
the PR intervals and RT intervals. The ruts still have
significant differences for normal and abnormal in our analysis.