生物光譜資訊分析及量測技術之優化

本研究主要目標在於開發一個低成本且可普及推廣的生物溶液光譜量測系統，並藉由訊號處理、樣品雜質過濾等方式有效提升光譜測量靈敏性及解析度。我們使用自行開發的系統量測及分析花青素、葉綠素、血紅素、油酸．．．等許多與每日生活息息相關的生物溶液，建立其吸收、螢光及拉曼光譜之資料庫。此外，我們也進一步將此量測系統應用至糞便潛血感測上。根據全血溶液的體積百分濃度量測結果，此開發系統可有效辨識潛血反應之感測靈敏度約為0.15 μg-Hb/ 1 ml-water。針對實際檢體進行分析，我們發現由此開發系統辨識的結果與現行糞便潛血檢測的免疫法(iFOBT)所得的結果有良好的對應。若能進一步開發演算法與所建立生物溶液之光譜資料庫進行快速比對，將有機會將此開發系統推廣至更全面的飲食成分分析。

關鍵字

血紅素；糞便潛血；吸收光譜；螢光光譜

並列摘要

The main goal of this research is to develop a low-cost measuring system which can be popularized to be a houschold apparatus for analyzing optical spectra of biological solutions. Some signal processing and impurity filtering approaches for the measured samples are studied to effectively enhance the sensitivity and resolution of the developed system in detecting the characteristic spectral signals for biological solution. A thorough database for the absorption, fluorescence, and Raman spectra of some common biological solutions, such as anthocyanin, chlorophyll, hemoglobin, oleic acid, that is closely related to our daily life has been built by the developed system. Furthermore, the developed system is further applied for the detection of fecal occult blood. According to the measurement of absorption spectrum for whole blood, the effective sensitivity of the developed system to identify the occult blood signal is about 0.15 μg-Hb/ 1 ml-water. We find that the results of identification for the fecal occult blood in real-patient sample by the proposed method are highly consistent with the results by the conventional immune method in medical examination. We believe it is greatly feasible to further integrate an efficient algorithm into the proposed system to realize the component analysis for the sampling solution by matching the spectral characteristic signals with the spectrum database for some common biosolutions.