人体血清中葡萄糖浓度的近红外光谱分析

Analysis of Glucose Concentration in the Blood Serum by Near-Infrared Spectroscopy

血糖的无创监测为糖尿病的诊断节省了大量的时间和试剂 ,更重要的是大大减轻了糖尿病患者的痛苦。对血清中葡萄糖浓度做近红外光谱分析可以探索实验方法 ,积累实验数据。本文利用近红外光谱透射测量血糖浓度的吸光度变化 ,选择包含有葡萄糖特征峰的谱段建立校正模型 ,采用基线校正、曲线平滑和二次求导的方法对光谱曲线作预处理。用偏最小二乘法 (PLS)建立回归模型。相关系数达到 0 .95 16 1,均方差达到 0 .36

Noninvasive measure of the glucose saves so much reagent and time for the diabetes mellitus diagnoses,especially decreasing patient's agony.The method is developing and the data is accumulated through the analysis of the glucose near-infrared spectrum.The glucose in the blood serum was determined by near-infrared technology.In order to decrease the error and increase the correlation coefficient,the calibrating model was set up for characteristic peak by the partial least square algorithm (PLS).The best correlative coefficient is 0.95161 with the RMSEC up to 0.361.