全人工髖關節置換術為骨科常見手術之一,鬆脫是其術後最常產生之併發症。本研究以希伯特黃轉換及快速傅立葉轉換進行訊號分析,並以正確性、靈敏度及有效性評估此二種方法對於鬆脫訊號的解析能力,另探討不同量測位置所擷取訊號對於診斷鬆脫的正確性的影響。 本研究以非侵入式音頻量測系統擷取人體髖關節振動訊號,量測位置為:正躺-前上腸骨棘、側躺時的前上腸骨棘及大轉子。將訊號分別以希伯特黃轉換及快速傅立葉轉換進行分析,並依據其諧波振幅大小進行鬆脫的判斷,將系統判斷結果與醫師臨床診斷進行比對,以正確性、靈敏度及有效性三個參數來評估此二種訊換轉換方法的效能,期能找出較適於分析振動訊號的方式。另關於量測位置對於鬆脫診斷的影響,本研究亦有完整的討論。 結果顯示,以希伯特黃轉換分析訊號的正確性與靈敏度均高於快速傅立葉轉換的分析結果,顯示希伯特黃轉換確能提高以訊號診斷鬆脫的效能。另在量測位置的討論,側躺-大轉子與前上腸骨棘所擷取訊號,經希伯特黃轉換後其正確性與靈敏度均超過0.8,明顯高於正躺-前上腸骨棘,且量測的位置對於偵測不同類型的鬆脫有所影響。 研究發現以振動訊號診斷鬆脫是值得發展的,且希伯特黃轉換對於振動訊號有較佳的結果呈現。此研究提供一新的分析方向供後人在探討人體生理訊號時之參考。
Total hip replacement arthroplasty (THR) is a common operation on Orthopedics. The most serious common complication of THR is loosening of the prosthetic stem and cup. This study used Hilbert Huang Transform (HHT) and Fast Fourier Transform (FFT) to analyse the vibrational signals of different posture and position, then compared the results by using three parameters: Accuracy, Sensitivity and Specificity. The study used non-invasive sonic vibration measurement system to measure the vibrational signal of different posture and position, (a) lie on the back-Anterior Superior iliac Spine (ASIS), (b) lie on the side-ASIS, (c)lie on the side-Greater trochanter, and analysed the vibrational signals by HHT and FFT. According to the signal amplitude of harmonic, the system would indicate that the prosthesis was loosening or secure, and then compared with clinical diagnosis based on the patient’s radiograph. The signal resolution abilities of HHT and FFT were evaluated by three parameters. Besides, the inferences of postures and positions on vibrational signal had been completed discussion in this study. The results showed that the HHT had better accuracy and sensitivity than FFT, and it meant that the HHT could raise the ability in diagnosing loosening of THA. Furthermore, the measurement of postures and positions would affect the accuracy in different kinds of loosening . Using vibrational signal to detect loosening of THA was worth developing, and the HHT was a proper method in analysing loosening signal of THA.