摘要 糖尿病患者血糖濃度量測的便利性與準確性相對其病情的控制有著相當直接的關係。透過非侵入無痛的量測不但可以減少病人面對扎針的痛苦與不適、防止扎針傷口所造成的細菌感染以及降低診斷的花費。 本文主要探討Photoacustic非侵入式量測葡萄糖的方法,並與近紅外線光譜吸收法、旋光法、散射法相比較,PA法在應用上更具敏感性。本研究最主要的目的是以Photoacoustic法進行in vitro葡萄糖溶液的濃度量測,分別就受影響葡萄糖濃度影響的參數進行實驗,找出相互間的關係。藉由量測膨脹係數、比熱與波速的in vitro的實驗,證明這些參數對於acoustic wave pressure的影響力。並比較各種非侵入式的血糖檢測技術當中選擇一種安全、準確度高且適用於人體的方法實際應用於糖尿病患者的血糖檢測。 由實驗結果得到,使用的激發光源為905nm波長的雷射二極體照射不同濃度的葡萄糖水溶液時,其相關性為葡萄糖水溶液濃度變化1%,PA的振幅變化約為4.5%。本文為PA法量測葡萄糖的初步研究,期待未來能應用至人體,以達非侵入式量測血糖的目的。
Abstract Diabetes is one of the most important chronic diseases associated with lifestyle. Determination of blood glucose level is a frequently occurring procedure in diabetes care. Collecting blood drops for chemical analysis is the most common method. It is liable to afflict a degree of pain and cause a skin injury. To eliminate these disadvantages, this thesis focuses on pulsed photoacoustic techniques, which have potential ability in non-invasive blood glucose measurement. Compared with optical absorption techniques, Optical rotation and Brillouin scattering techniques, photoacoustic methods offer the advantage of higher detection sensitivity, as the PA signal is influenced not only by the optical absorption coefficient, but also by other physical parameters including thermal expansion, specific heat and acoustic velocity. The measurements showed that the expansion coefficient, specific heat and acoustic velocity change, respectively, in response to the change in glucose concentration and proved the sum effect of these parameters to photoacoustic signal. In this thesis, PA amplitudes were first deduced on the basis of the thermodynamic principle, which correspond very well with results deduced from the wave equation of thermal generation. At the wavelength of 905 nm, the measured glucose detection sensitivity in water solution, about 4.5% for a 1% change in glucose concentration. It is preliminary study for blood glucose determinate. The anticipation future will be able to apply to the human body measurement, to achieve the goal for non- invasive diagnosis.