本研究量測脊髓損傷病患於傾斜床訓練過程中,從正常生理狀態到發生起坐性暈眩現象之連續心電圖訊號,以希伯特黃轉換(Hilbert Huang Transform, HHT)與快速傅立葉轉換(Fast Fourier Transform, FFT)分析病患之心率變異高低頻譜,作訊號能量、解析度特性之比較。結果發現,希伯特黃轉換理論可改善傳統訊號分析方式,使訊號能呈現較佳的訊息與特性。 另一方面,由於過去學者於研究頸脊髓損傷病患之起坐性暈眩現象,未有可靠評估暈眩的定性描述或定量指標,且傾斜角度的刺激及病患痊癒的程度亦有可能影響量測之生理值。因此,本研究採用起坐性暈眩假說來完整描述病患起坐性暈眩與心率變異度之關係;其心率變異分析係利用希伯特黃轉換與快速傅立葉轉換加以比較,以驗證此兩種心率變異訊號處理方式是否符合此暈眩假說。 研究結果顯示:於訊號能量比較方面,以HHT分析之LF/HF頻譜有更顯著之特徵變化表現;於解析度比較方面,FFT於短時間視窗(30sec)下,無法解析訊號真正特性;於驗證起坐性暈眩假說方面,HHT較FFT符合此暈眩假說之定義。 本研究經過實驗數據的驗證,結果顯示HHT確實在應用上有其優勢;其結果可應用在臨床上,以提供醫療人員處理暫態、非週期性與非平穩性之心率變異訊號分析及提高病患復健治療的效率。
This research investigated the spectrum analyses of heart rate variability (HRV) in spinal cord injury (SCI) patients from normal condition to orthostatic fainting during tilting table training, and compared Hilbert Huang Transform (HHT) with Fast Fourier Transform (FFT) in the qualities of power spectrum and resolution. We found that HHT would be a better signal processing method than FFT for HRV analyses. Furthermore, many investigators studied orthostatic fainting of SCI patients, but there was no dependable method of qualitative descriptions and quantitative targets. Therefore, this research finally adopted the hypothesis of orthostatic fainting that could describe the relationship between orthostatic fainting and HRV completely. The spectrum analyses of HRV that fit in with the hypothesis of orthostatic fainting were also compared HHT with FFT. The results could be divided into three parts. With regard to power spectrum, the trend of LF/HF spectrum analyzed by HHT was more conspicuous than FFT. In regard to resolution, FFT could not differentiate the signal properties in a short time. With reference to hypothesis of orthostatic fainting, HHT tallied with the definitions of hypothesis completely. This research originally adopted Hilbert Huang Transform to analyze heart rate variability, and the results showed that Fast Fourier Transform had defects and limits in applications. We suggest that HHT can be applied to analyze heart rate variability better than FFT in the future.