近年使用非侵入式傳遞速度(Pulse wave velocity, PWV) 量測血壓的研究受到學術界的重視，研究人員使用心電圖(ECG)與光體積血容積(Photoplethysmograph,PPG)的時間差，或是光體積血容積（PPG）在手指到腳趾、或手指到手腕的脈波傳導速率，來進行收縮壓(Systolic blood pressure,SBP)與舒張壓(Diastolic blood pressure, DBP)的估測。然而，使用心電圖必須量測身體的左右兩端，而使用光體積血容積（PPG）需要量測兩個身體的遠端，這些都造成使用上的不便利，也不適合穿戴式裝置使用。 因此，本論文使用量測單手的光體積血容積（PPG）時間差進行血壓估測，藉由量測前臂與手腕的量測點，測量脈波傳遞時間(Pulse transit time, PTT)以估測血壓波。我們量測12位正常人(24±2.5歲)、以及15位患有高血壓及糖尿病病人(55±6.5歲)。為了誘發收縮壓的提升，我們利用跑跑步機五到十分鐘，使我們的收縮壓分布有高低之分，並發現運動後血壓上升所引起的脈波傳遞時間，符合脈波傳遞時間與壓力成反比的趨勢，經過與市售血壓機所量測的收縮壓做線性迴歸分析，並推得正常人在初始狀態誤差 3.72±3.38%，病人為12.28±11.77%。 ;The utilization of noninvasive pulse wave velocity (PWV) to estimate blood pressure has drawn great attention in recent years. Researchers measure the time difference between feature points of electrocardiography (ECG) and photoplethysmography (PPG) to continuously estimate systolic blood pressure (SBP) and diastolic blood pressure (DBP). However, the measurement of ECG require electrode placements across left and right sides of human body, and the time difference of PPGs was detected by placing two PPG signals measured from two distal position on human body. Both of the two aforementioned methods are inconvenient to be used and not suitable to be realized on wearable devices. Accordingly, this study aims to develop a noninvasive blood pressure measurement method based on measuring two PPG signals from a single hand. By detecting pulse transit time (PTT) between feature points of two green light PPG signals, one measured from forearm and the other from wrist, the model of systolic blood pressure versus pulse wave velocity (PWV) is established. In this study, we have measured twelve normal subjects (mean age = 24±2.5 years-old) and fifteen patients, who are suffering from hypertension and diabetes (mean age = 55±6.5 years-old). Each subject was requested to run on a treadmill for five-to-ten minutes. We found the difference between before-exercise and after-exercise PTT was inversely proportional to the measured systolic blood pressure (SBP), and blood pressure values were estimated by means of creating a linear regression model. The estimated error of SBP in normal people was 3.72±3.38%, and the SBP error in hypertension patient was 12.28±11.77%
