摘要
光學式脈動血氧濃度計是一種非侵入性與即時性的量測儀器，在手術房、重症看護以及居家照護的監測上，血氧濃度計是計算紅光與近紅外光的脈動訊號的比例值來由校正曲線上取得對應的血氧濃度值。血氧濃度計在低灌流量與低血氧濃度時的準確度問題一直是對研究人員的強大挑戰，確保量測準確度的校正方式主要是仰賴大量人體測量數據中來產生適合之校正曲線。本研究以比爾-朗伯定律為基礎分析血氧濃度校正曲線的形成及影響準確度的變因，同時也分析量化解析度對準確度的影響，並以模擬光脈動訊號套入離散飽和轉換與R值回歸方程式來分析兩者的差異。 由模擬結果可知，在PI為20%時使用8 bits的解析度即可符合FDA對準確度的要求，但需使用9 bits以上方能符合ISO-80601-2-61的新規範；而PI=0.3%的訊號則需分別使用13 bits及15 bits以上。在血氧濃度90~100%時，依最大誤差評比主要的影響因素依序為紅光中心波長偏移、LED寬光譜光源的分布、紅光與近紅外光的路徑長比例。
並列摘要
Pulse oximeter is a non-invasive and real-time measurement instrument. It is widely used in operating room and critical care units, as well as for home care monitoring. Pulse oximeter calculates the ratio of pulsatile signals of red and near Infrared light to obtain the oxygen saturation of blood from its calibration curve. The accuracy of pulse oximeter at low perfusion rate and low oxygen saturation has been a great challenge to researchers. The accuracy of pulse oxygen saturation relies on a large number of measurements on human studies to obtain a calibration curve. In this study, the Beer-Lambert’s law was used to derive the calibration curve of pulse oximeter and to analyze the factors that affects the accuracy of measurement. The influence of the resolution of analog-to-digital on accuracy was also analyzed. Finally, simulated pulse oximetric signals are used to compare the calculation of oxygen saturation using calibration curve and discrete saturation transformation. The simulation results show that 8 bits resolution of A/D conversion is enough to meet the FDA requirements of accuracy when the perfusion rate is 20%, but 9 bits resolution is needed to meet the new ISO-80601-2-61 standard. if the perfusion rate is as law as 0.3%, resolution of 13 bits and 15 bits are required, respectively. When oxygen saturation is in the range between 90 and100%, the order of factors that contributes to the error in measurement are wavelength shift of red light, spectral width of LED light source, and the path length ratio of red and near infrared light.