以龐加萊及相空間重構法分析睡眠呼吸中止症心跳特徵

本研究利用了渾沌理論中的龐加萊以及相空間重構法，分析受試者的睡眠心跳變異率，找出睡眠呼吸中止症的心跳特徵。龐加萊方法是將心跳變異率的x(i)作為橫軸，x(i+1)作為縱軸，以逆時針旋轉後，成為與軸，接著計算座標值與及軸之標準偏差量SD1與SD2，進行後續分析。相空間重構是將心跳變異率x(i)先藉由平均互訊息法決定延遲時間，接著由假鄰近點方法決定嵌入維度後，計算相空間李亞普諾指數，進行後續分析。研究中有十七位健康正常人與十九位睡眠呼吸中止症患者，共三十六位受試者進行睡眠檢查，取得整晚所紀錄的心電圖以及睡眠技師所判讀的睡眠階段後，將各睡眠階段心電圖中的所有R波時間取出，形成R-R時間差的心跳變異率序列。龐加萊分析結果發現，睡眠呼吸中止症病患整晚的交感神經指標SD1/SD2顯著較大(p=0.025)。另外，也發現病患在睡眠第一期(p=0.022)、睡眠第二期(p= 0.019)、睡眠第三期(p=0.002)、以及快速眼動期(p<0.001)交感神經指標SD1/SD2都顯著的較健康人大，這樣的結果顯示呼吸中止症患者在上述睡眠期間交感神經都比健康人的緊張。病患的副交感神經指標SD1*SD2在快速眼動期時，比健康人要來小，且有顯著差異(p<0.001)，顯示在快速眼動期副交感神經都作用較弱。相空間重構結果發現健康人與呼吸中止症患者在整晚的心跳變異率的嵌入維度相差不大，但分析相空間軌跡所計算的李亞普諾指數後發現，病患整晚的心跳變異率李亞普諾指數顯著較健康人小(p=0.009)，另外，呼吸中止症患者在睡眠清醒期(p=0.002)、睡眠第一期(p=0.018)、睡眠第二期(p<0.001)以及快速眼動期(p=0.002)中有比健康人低的李亞普諾指數，顯示健康人有較為渾沌的心跳。比起需要量測口、鼻氣流的傳統方法，本研究提出只要心電訊號，就可成為初步篩檢睡眠呼吸中止症的簡易輔助工具

關鍵字

睡眠呼吸中止症；心跳變異率；睡眠階段；龐加萊；相空間重構；李亞普諾指數

並列摘要

This study develops ECG markers of sleep apnea syndrome based on all-night sleep ECG. First of all, the sleep ECG have been recorded for 17 normal controls and 19 sleep apnea patients. Second, ECG is transformed to R-R intervals and named HRV. Third, using the Poincare Map and Reconstructed Phase Space to analyze HRV in sleep. Using Poincare Map, the plot of x(i) and x(i+1) of HRV, and oriented with the line-of-identity are denoted by and ,then the dispersion of the points around and are measured by the standard deviation denoted by SD1 and SD2. Using Reconstructed Phase Space, determine the delay time by using Average Mutual Information and then followed by using nearest neighbor to determine the invading dimension, and calculating the Lyapunov exponent of Reconstructed Phase Space. Poincare map’s results shown that, in all night HRV, two groups are significant difference in P value (p=0.025) in SD1/SD2, and in sleep stage 1(p=0.022), 2(p=0.019), 3(p=0.002) and REM(p<0.001), are also significant difference in P value in SD1/SD2, which represents sleep apnea group’s sympathetic nerve level are higher. On the other hand, two groups are significant difference in P value in sleep stage REM (p<0.001) in SD1*SD2, which represents the parasympathetic nerve. This means control group are more relax than sleep apnea group in this stages. Reconstructed Phase Space’s results shown that the invading dimensions are no difference between control group and sleep apnea group, but the Lyapunov exponents between two groups are significant difference in P value (P<0.009), and in sleep stage wake(p=0.002), 1(p=0.018), 2(p<0.001) and REM(p=0.002) are also significant difference in P value. This means control group’s HRV have higher level of chaos. This method is advantageous, instead of nasal flow and oral flow, only ECG channel measurement is required to identify sleep apnea syndrome, and possesses a superior extension for further research.

並列關鍵字

Sleep Apnea Syndrome ； Heart Rate Variability ； Sleep Stages ； Poincare map ； Reconstructed Phase Space ； Lyapunov Exponent

參考文獻

Abarbanel, Henry DI, and Jerry P. Gollub. "Analysis of observed chaotic data." Physics Today 49 (1996): 86.

Abarbanel, Henry DI, and Matthew B. Kennel. "Local false nearest neighbors and dynamical dimensions from observed chaotic data." Physical Review E 47.5 (1993): 3057.

Al-Angari, Haitham M., and Alan V. Sahakian. "Use of sample entropy approach to study heart rate variability in obstructive sleep apnea syndrome." Biomedical Engineering, IEEE Transactions on 54.10 (2007): 1900-1904.

Baharav, A., Kotagal, S., Rubin, B. K., Pratt, J., & Akselrod, S. (1999). Autonomic cardiovascular control in children with obstructive sleep apnea. Clinical Autonomic Research, 9(6), 345-351.

Brennan, M., M. Palaniswami, and P. Kamen. "Do existing measures of Poincare plot geometry reflect nonlinear features of heart rate variability?." Biomedical Engineering, IEEE Transactions on 48.11 (2001): 1342-1347.

國際替代計量

以龐加萊及相空間重構法分析睡眠呼吸中止症心跳特徵

全文下載

主題瀏覽